O'Donnell Et Al. 2014

YFGBI 2750 No.
of Pages 14, Model 5G

15 November 2014
Fungal Genetics and Biology xxx (2014) xxx–xxx

1
Contents lists available at ScienceDirect
Fungal Genetics and Biology

journal homepage: www.elsevier.com/locate/yfgbi
5
6
3 Discordant phylogenies suggest repeated host shifts

4 in the Fusarium–Euwallacea ambrosia beetle mutualism
7 Q1 Kerry O’Donnell a,⇑, Stacy Sink a, Ran Libeskind-Hadas b, Jiri Hulcr c, Matthew T. Kasson d, Randy C. Ploetz e,
8 Joshua L. Konkol e, Jill N. Ploetz e, Daniel Carrillo e, Alina Campbell e, Rita E. Duncan e,
9 Pradeepa N.H. Liyanage f, Akif Eskalen g, Francis Na g, David M. Geiser h, Craig Bateman c,
10 Stanley Freeman i, Zvi Mendel i, Michal Sharon i, Takayuki Aoki j, Allard A. Cossé k, Alejandro P. Rooney k
11 a
Bacterial Foodborne Pathogens and Mycology Research Unit, National Center for Agricultural Utilization Research, US Department of Agriculture, Agricultural Research Service,
12 1815 North University Street, Peoria, IL 61604, USA
13 b
Department of Computer Science, Harvey Mudd College, Claremont, CA 91711, USA
14 c
School of Forest Resources and Conservation, University of Florida, Gainesville, FL 32611, USA
15 d
Division of Plant and Soil Sciences, West Virginia University, Morgantown, WV 26506, USA
16 e
Tropical Research and Education Center, University of Florida, Homestead, FL 33031, USA
17 f
Tea Research Institute of Sri Lanka, St. Coombs, Talawakelle, Sri Lanka
18 g
Department of Plant Pathology and Microbiology, University of California, Riverside, CA 92521, USA
19 h
Department of Plant Pathology and Environmental Microbiology, Pennsylvania State University, University Park, PA 16802, USA
20 i
Institute of Plant Protection, ARO, The Volcani Center, Bet Dagan 50250, Israel
21 j
National Institute of Agrobiological Sciences, Genetic Resources Center, 2-1-3 Kannondai, Tsukuba, Ibaraki 305-8602, Japan
22 k
Crop Bioprotection Research Unit, National Center for Agricultural Utilization Research, US Department of Agriculture, Agricultural Research Service, 1815 North University Street,
23 Peoria, IL 61604, USA
24
25
a r t i c l e i n f o a b s t r a c t
2
4 7
0
28 Q2 Article history: The mutualism between xyleborine beetles in the genus Euwallacea (Coleoptera: Curculionidae: Scolyti- 41
29 Received 28 August 2014 nae) and members of the Ambrosia Fusarium Clade (AFC) represents one of 11 known evolutionary origins 42
30 Accepted 27 October 2014 of fungiculture by ambrosia beetles. Female Euwallacea beetles transport fusarial symbionts in paired 43
31 Available online xxxx
mandibular mycangia from their natal gallery to woody hosts where they are cultivated in galleries as a 44
source of food. Native to Asia, several exotic Euwallacea species were introduced into the United States 45
32 Keywords: and Israel within the past two decades and they now threaten urban landscapes, forests and avocado 46
33 Cophylogeny
production. To assess species limits and to date the evolutionary diversification of the mutualists, we 47
34 Fungiculture
35 Hybrid introgression
reconstructed the evolutionary histories of key representatives of the Fusarium and Euwallacea clades 48
36 Molecular phylogenetics using maximum parsimony and maximum likelihood methods. Twelve species-level lineages, termed 49
37 Phylogenetic species AF 1–12, were identified within the monophyletic Ambrosia Fusarium Clade (ACF) and seven among the 50
38 Symbiosis Fusarium-farming Euwallacea. Bayesian diversification-time estimates placed the origin of the Euwallacea 51
39 –Fusarium mutualism near the Oligocene–Miocene boundary 19–24 Mya. Most Euwallacea spp. appear to 52
be associated with one species of Fusarium, but two species farmed two closely related fusaria. Euwallacea 53
sp. #2 in Miami-Dade County, Florida cultivated Fusarium spp. AF-6 and AF-8 on avocado, and Euwallacea 54
sp. #4 farmed Fusarium ambrosium AF-1 and Fusarium sp. AF-11 on Chinese tea in Sri Lanka. Cophyloge- 55
netic analyses indicated that the Euwallacea and Fusarium phylogenies were largely incongruent, appar- 56
ently due to the beetles switching fusarial symbionts (i.e., host shifts) at least five times during the 57
evolution of this mutualism. Three cospeciation events between Euwallacea and their AFC symbionts were 58
detected, but randomization tests failed to reject the null hypothesis that the putative parallel cladogen- 59
esis is a stochastic pattern. Lastly, two collections of Euwallacea sp. #2 from Miami-Dade County, Florida 60
shared an identical cytochrome oxidase subunit 1 (CO1) allele with Euwallacea validus, suggesting 61
introgressive hybridization between these species and/or pseudogenous nature of this marker. Results 62
of the present study highlight the importance of understanding the potential for and frequency of 63
host-switching between Euwallacea and members of the AFC, and that these shifts may bring together 64
more aggressive and virulent combinations of these invasive mutualists. 65
Published by Elsevier Inc. 66
68
67
69
⇑ Corresponding author.
E-mail address: kerry.odonnell@ars.usda.gov (K. O’Donnell).
http://dx.doi.org/10.1016/j.fgb.2014.10.014
1087-1845/Published by Elsevier Inc.
Please cite this article in press as: O’Donnell, K., et al. Discordant phylogenies suggest repeated host shifts in the Fusarium–Euwallacea ambrosia beetle
mutualism. Fungal Genet. Biol. (2014), http://dx.doi.org/10.1016/j.fgb.2014.10.014
YFGBI 2750 No. of Pages 14, Model 5G
15 November 2014
2 K. O’Donnell et al. / Fungal Genetics and Biology xxx (2014) xxx–xxx
70 1. Introduction death in both young and mature trees (Mendel et al., 2012). Fusar- 134
ium dieback of avocado caused little concern when it was first 135
71 The Fusarium–Euwallacea (Coleoptera: Curculionidae: Scolyti- observed in Israel in 2005, but this disease is now responsible for 136
72 nae) mutualism represents one of 11 known evolutionary origins serious damage mainly on avocado, box elder (Acer negundo L.), 137
73 of fungiculture by scolytine ambrosia beetles (Kasson et al., 2013). castor bean (Ricinus communis L.), and English oak (Quercus robur 138
74 Euwallacea is a genus of over 40 species within the largest tribe of L.) in that country (Mendel et al., 2012). These four hosts are also 139
75 scolytine beetles, the Xyleborini, which contains nearly 1200 among the 32 preferred by the PSHB in California (http://eskalen- 140
76 described species (Hulcr et al., 2015). All Xyleborini farm ambrosia lab.ucr.edu/avocado.edu.html), based on its ability to produce via- 141
77 fungi – symbionts that concentrate nutrients from their woody ble offspring on them (Eskalen et al., 2013). Although the PSHB was 142
78 hosts. Most Xyleborini are associated with fungi from Ophiostoma- originally discovered in the Los Angeles Basin in 2003, the disease 143
79 tales to Microascales (Beaver, 1989). Euwallacea is the only genus was not observed on avocado in California until 2012 (Eskalen 144
80 known to cultivate ambrosial fusaria, sometimes in addition to et al., 2012) where it damaged the above hosts, as well as California 145
81 microascalean and ophiostomatalean fungi. The fusaria associated coast live oak (Quercus agrifolia Née), California sycamore (Platanus 146
82 with Euwallacea form a monophyletic group (Ambrosia Fusarium racemosa Nutt.), and many other tree species (Eskalen et al., 2013). 147
83 Clade, or AFC) within the Fusarium solani species complex, originally Conspicuous white eruptions of mannoheptulose and perseitol 148
84 reported to encompass 11 species lineages, nine of them (AF 1–9) referred to as ‘sugar volcanoes’ develop on affected avocado trees 149
85 with known associations with ambrosia beetles (Kasson et al., (see Fig. 2 in Mendel et al., 2012). These sugars are produced on 150
86 2013). Published diversification-time estimates of the Xyleborini branches and trunks of avocado in response to beetle entry and 151
87 dated their evolutionary radiation to the early Miocene 21 Mya subsequent necrosis caused by F. euwallaceae (Freeman et al., 152
88 (Jordal and Cognato, 2012), similar to the inferred divergence time 2013; Mendel et al., 2012). 153
89 of the AFC (Kasson et al., 2013), suggesting that Euwallacea’s associ- Besides two separate infestations of E. fornicatus-like beetles in 154
90 ation with Fusarium, originated early after the origin of the whole Southern California (i.e., San Diego County and Los Angeles and sur- 155
91 fungus-farming tribe. Because there are no inferred evolutionary rounding counties), and a third infestation in Miami-Dade County, 156
92 reversals to phloem-feeding, all of the fungus-farming beetles are Florida, two other exotic species of Euwallacea originally from Asia 157
93 thought to be obligate mutualists (Farrell et al., 2001). Female are now established in the US: Euwallacea validus Eichhoff and 158
94 Euwallacea carry AFC mutualists in specialized mandibular cavities interjectus interjectus Blandford. E. validus is a temperate species 159
95 located within their mandibles termed mycangia (Beaver, 1989). with broad tree host range (Rabaglia et al., 2006); a population 160
96 During the past two decades, several Euwallacea have been increase has been documented on the invasive Asian tree of heaven 161
97 introduced from their native areas in Asia into Israel, Central Amer- (Ailanthus altissima (Mill.) Swingle) symptomatic for Verticillium 162
98 ica and at least five different locations within the United States wilt in eastern North America (Kasson et al., 2014). E. interjectus is 163
99 (Carrillo et al., 2012; Eskalen and Stouthamer, 2012; Haack, a subtropical species spreading throughout the Southeastern US 164
100 2006; Kasson et al., 2013; Kirkendall and Ødegaard, 2007; Ploetz (Atkinson, 2014). Mass attack of this species has been documented 165
101 et al., 2013), presumably on infested wood packaging or plant on box elder in Gainesville, Florida (Kasson et al., 2013). 166
102 material (Wingfield et al., 2010). The best-known example is the One of the most challenging aspects of research on the invasive 167
103 tea shot hole borer (TSHB), Euwallacea fornicatus Eichhoff, and its Euwallacea spp. and their Fusarium symbionts has been defining 168
104 symbiont Fusarium ambrosium (Gadd and Loos) Agnihothr. and taxonomic units, particularly species. The beetles are haplo-diploid 169
105 Nirenberg. In their native region in southern Asia, these symbionts inbreeders, which means that the great majority of matings occur 170
106 mostly colonize dead or declining species from at least 48 plant within a family between a single haploid male and his diploid sis- 171
107 Q3 families (Danthanarayana, 1968; Hulcr et al., 2007), but they are ters (Kirkendall, 1993). Consequently, in the inbred Xyleborini, 172
108 also known as destructive pests of several economically important morphological characters change in a gradual fashion across 173
109 woody plants, including Chinese tea (Camellia sinensis) (L.) Kuntze), clades, instead of being relatively uniform within species and dif- 174
110 avocado (Persea americana Mill.), citrus (Citrus spp.) and cacao ferent between species, as in most animals. Thus, applying a phy- 175
111 (Theobroma cacao L.), where they can cause extensive dieback logenetic species concept within Euwallacea seems warranted 176
112 and even death (Brayford, 1987). given that the utility of traditional species concepts based on mor- 177
113 Recently, a new disease of avocado in California and Israel, phology or mating barriers remains to be determined. This is 178
114 Fusarium dieback, was initially reported to be associated with important in groups such as the E. fornicatus complex, where clades 179
115 E. fornicatus (Eskalen et al., 2012; Eskalen and Stouthamer, 2012; that are uniform morphologically may be distinct phylogenetically 180
116 Mendel et al., 2012). However, because preliminary molecular and ecologically, including different levels of aggression towards 181
117 systematic data suggested the beetle in California and Israel was host trees. In this work, several of those populations are included, 182
118 genetically distinct from E. fornicatus in Sri Lanka, it was given and are denoted only as Euwallacea sp., rather than E. fornicatus, in 183
119 the common name polyphagous shot hole borer (PSHB) to distin- the absence of robust species-level molecular phylogenetic and 184
120 guish it from the TSHB (Eskalen et al., 2013). Results of an exten- type studies to determine what species represents the TSHB. 185
121 sive survey conducted in the aforementioned study revealed that Similarly, the AFC fusaria have also evaded taxonomic treat- 186
122 the PSHB had attacked over 200 woody hosts representing 58 plant ment of species until recently (Freeman et al., 2013; Kasson 187
123 families in the Huntington Botanical Garden and Los Angeles et al., 2013). They comprise multiple morphologically cryptic spe- 188
124 Arboretum in the San Gabriel Valley of Southern California cies that are only known to reproduce asexually. In our previous 189
125 (Eskalen et al., 2013). In contrast to the TSHB that cultivates F. molecular phylogenetic study of the AFC, F. ambrosium was the 190
126 ambrosium on tea in India and Sri Lanka, the Euwallacea sp. in Israel only species that had been described formally (Gadd and Loos, 191
127 and California farms a closely related species recently described as 1947; Nirenberg, 1990). Based on morphological differences from 192
128 Fusarium euwallaceae (Freeman et al., 2013), which causes wilt and F. ambrosium and phylogenetics, F. euwallaceae was described for 193
129 dieback of avocado and diverse landscape trees in concert with AF-2, the AFC lineage associated with the Israeli and Los Angeles 194
130 mass attack by the beetle (Eskalen et al., 2013). In this disease, Basin outbreaks (Eskalen et al., 2013; Freeman et al., 2013). 195
131 the fungus moves from the beetle galleries along the xylem To clarify the identity and evolutionary history of the globally 196
132 elements and adjacent cells, producing a brown discoloration and invasive representatives of the Euwallacea–Fusarium symbiosis, 197
133 necrosis that often results in loss of yield, dieback and ultimately the present study was conducted to (i) infer genealogically exclusive 198
15 November 2014
K. O’Donnell et al. / Fungal Genetics and Biology xxx (2014) xxx–xxx 3
199 species boundaries (Taylor et al., 2000) by analyzing a 4-gene data- Supplementary Table S5). Relationships of the samples were 259
200 set for the AFC and a 6-gene dataset for the Euwallacea that cultivate inferred using a combined maximum parsimony phylogenetic 260
201 them, (ii) date the evolutionary origin and diversification of the analysis (MP, Swofford, 2003) of the 28S rDNA and CO1 sequences. 261
202 Fusarium–Euwallacea mutualism, and (iii) conduct cophylogenetic Based on this phylogeny, 21 Euwallacea individuals from the three 262
203 analyses to determine the relative roles of cospeciation and/or targeted species groups (E. fornicatus, E. validus and E. interjectus) 263
204 host-shift speciation in the evolution of this increasingly important were selected to represent the range of phylogenetic diversity, 264
205 mutualism. and they were additionally typed using portions of four more genes 265
(Table 1): translation elongation factor (EF-1a, 373 bp alignment), 266
arginine kinase (ArgK, 669 bp alignment), CAD protein (CAD, 267
206 2. Materials and methods
469 bp alignment), and mitochondrial 16S rDNA (16S rDNA, 268
420 bp alignment). PCR amplification and DNA sequencing fol- 269
207 2.1. Taxon sampling
lowed published protocols (Dole et al., 2010; Simon et al., 1994; Q4 270
Supplementary Table S3). 271
208 In contrast to males that lack mycangia and are flightless,
Because two of the 103 Euwallacea individuals we typed 272
209 female Euwallacea possess paired mandibular mycangia within
appeared to represent E. validus – Euwallacea sp. #2 interspecific 273
210 which they transport their AFC symbionts (Kasson et al., 2013).
hybrids (Fig. 2; RP-30 and RP-31), the following steps were followed 274
211 Thus, to elucidate the fidelity of the Euwallacea–Fusarium mutual-
to rule out the possibility of technical error and cross-contamination 275
212 ism, 103 adult female Euwallacea ambrosia beetles were collected
of the samples with E. validus DNA. DNA was extracted separately 276
213 at diverse localities within the United States, Israel, Sri Lanka,
from the two beetles using dedicated disposable pestles and PCR 277
214 Papua New Guinea and Australia for DNA typing (Supplementary
amplification and DNA sequencing of the nuclear and mitochondrial 278
215 Table S1). Some beetles were obtained using sticky or funnel traps,
genes were repeated on three separate occasions. It is worth noting 279
216 but the majority were recovered directly from host trees. Once col-
that the two putative E. validus – Euwallacea sp. #2 hybrid beetles 280
217 lected, beetles were either placed immediately in 95–100% ethanol
were received in separate tubes filled with 100% ethanol from R. Plo- 281
218 or stored at 20 °C in the laboratory before shipping to NCAUR,
etz’s lab where E. validus has never been processed. PCR primer pairs 282
219 Peoria, Illinois for DNA typing, or they were transported live to
specific to the nuclear 28S rDNA, mitochondrial 16S rDNA and CO1 283
220 the laboratory. Fusaria were isolated from live-collected beetles
of Euwallacea sp. #2 and E. validus were designed and tested on 284
221 by dilution plating whole beetles or heads that had been surface
genomic DNA from several E. validus and Euwallacea sp. #2 individ- 285
222 sterilized in 70% ethanol for 15 s and macerated in sterile water
uals, and the two putatively E. validus – Euwallacea sp. #2 hybrids. 286
223 using a Tenbroeck tissue grinder (PYREXÒ No. 7727-07, Corning,
We reasoned that if the two putative hybrid samples had been 287
224 NY). Macerates of each collection were serially diluted one, 10
contaminated with E. validus DNA, as opposed to only having an 288
225 and 100 times, and aliquots of the dilutions were spread on half-
introgressed CO1 from E. validus, then the 28S and 16S rDNA of 289
226 strength potato dextrose agar (½ PDA) amended with 100 lg ml1
E. validus would have also amplified from these samples. The PCR 290
227 of streptomycin sulfate (½PDA + SS, Sigma–Aldrich, St. Louis, MO)
assay specific to CO1 of Euwallacea sp. #2 also allowed us to assess 291
228 so that the number of viable AFC propagules could be quantified
whether the putative interspecific hybrids possessed an ortholo- 292
229 (Kasson et al., 2013). Once a head was removed with a sterile scal-
gous allele. The remaining combinations of primers and species 293
230 pel under a dissecting microscope, each thorax + abdomen was
were conducted as a positive control for the PCR. 294
231 stored in a separate 1.5 mL Eppendorf tube containing 95–100%
Fusaria were initially typed using the intron rich 50 end of EF-1a 295
232 ethanol for subsequent DNA typing. The 91 fusaria that were
(687 bp alignment), which resolves all known species within the 296
233 DNA typed in the present study (Supplementary Table S2) are
AFC (Kasson et al., 2013). Subsequently, for putatively novel spe- 297
234 stored in 15% glycerol at -80 °C in the laboratory of K. O’Donnell
cies-level lineages within the AFC, portions of the following three 298
235 at NCAUR, Peoria, IL. Fusaria that were analyzed phylogenetically
nuclear genes were sequenced: rDNA internal transcribed spacer 299
236 using multilocus DNA sequence data are also available upon
region (ITS rDNA) and 28S rDNA D1/D2 domains (1004 bp align- 300
237 request from the ARS Culture Collection (NRRL), National Center
ment), DNA-directed RNA polymerase II largest (RPB1) subunit 301
238 for Agricultural Utilization Research (NCAUR), Peoria, Illinois USA
(1588 bp alignment), and DNA-directed RNA polymerase II second 302
239 (http://nrrl.ncaur.usda.gov/TheCollection/Requests.html). Other
largest (RPB2) subunit (1635 bp alignment) (Supplementary 303
240 fungal symbionts were recovered from Euwallacea, including Raffa-
Table S4). Multilocus DNA sequence data for 18 of the 25 fusaria 304
241 elea and Graphium from E. validus (Kasson et al., 2013), but they
included in the present study (Table 2) were published previously 305
242 were not included in this study because their presence was incon-
(Kasson et al., 2013). 306
243 sistent in the various life stages examined.
As a prerequisite for our cophylogenetic analyses of the Fusar- 307
ium–Euwallacea mutualism, phylogenetic species of the beetles 308
244 2.2. Gene sampling and the fungi were delimited employing the operational criteria 309
of genealogical concordance (GCPSR sensu Taylor et al., 2000) 310
245 Total genomic DNA was isolated using an acetyl trimethyl- and non-discordance (Dettman et al., 2003). Phylogenetic species 311
246 ammonium bromide (CTAB; Sigma–Aldrich, St. Louis, MO) protocol were recognized if bootstrapping of one or more of the individual 312
247 from a whole female Euwallacea beetle or the thorax + abdominal partitions and the combined dataset supported their genealogical 313
248 region and from approximately 100 mg of freeze-dried mycelium exclusivity at P70%, and their monophyly was not contradicted 314
249 of each AFC isolate that had been pulverized with a sterile pipette by bootstrap analyses of any of the individual gene partitions. 315
250 tip (O’Donnell et al., 1998). The generation of phylogenetic markers DNA sequences generated in the present study have been depos- 316
251 for both Euwallacea and AFC proceeded in two steps. First, the 103 ited in GenBank under accession numbers KM406624–KM406751 Q5 317
252 Euwallacea individuals collected were typed using DNA sequence and the alignments were deposited in TreeBASE as Tr77344, 318
253 data from two markers: a portion of the nuclear large subunit Tr77755, Tr77343, and Tr77754. 319
254 28S rDNA D2/D3 region (1004 bp alignment) and mitochondrial
255 cytochrome oxidase subunit 1 (CO1) gene (689 bp alignment). 2.3. Phylogenetic analysis and diversification-time estimates 320
256 DNA sequence data from three additional Xyleborini ambrosia bee-
257 tle species (Cognato et al., 2011) were downloaded from GenBank The individual and combined Euwallacea and Fusarium datasets 321
258 for inclusion in the phylogenetic analyses of Euwallacea (Fig. 1, were analyzed using maximum parsimony (MP, Swofford, 2002) 322
15 November 2014
4 genes Fusarium Euwallacea 6 genes

4914 bp 3630 bp
100 F .euwallaceae [AF-2] 499 PIC
388 PIC 0.6 Mya
9 MPTs 0.02−1.5 Israel / CA E. sp. #2 100 180 MPTs
81 951 steps
566 steps Miami-FL
CI = 0.78 99/84 CI = 0.79
F.sp. [AF-3]
RI = 0.87 1.2 Mya
RI = 0.90
0.07−2.7 Gainesville-FL
10 20
steps 1.5 Mya steps

99 0.5−3.1
F.sp. [AF-4] E. sp. #4
72
PA-OH-VA-MD Sri Lanka
1.6 Mya 97/83 2.3 Mya
0.11−3.5 100 0.8−4.3
100
F.sp. [AF-12] E. sp. #3 100
San Diego-CA
Australia 2.5 Mya
2.4 Mya 1.0−4.9
0.18−4.9 E. sp. #6
100 F.sp. [AF-5] E. ‘fornicatus’ PNG
3.1 Mya Malaysia Euwallacea

100
0.23−6.3 fornicatus
100 Clade
E. sp. #1
F.sp. [AF-11] 6.5 Mya
2.3 Mya
Sri Lanka 100 2.8−11.2
0.07−4.8 Israel / CA
F.sp. [AF-10]
Singapore 100
3.8 Mya
0.38−7.7 100 80/62
F. ambrosium [AF-1]
E. sp. #5 3.2 Mya
100 100
India / Sri Lanka 1.1−6.4
Clade B San Diego-CA
10.3 Mya
4.9 Mya 4.9−16.9
0.7−10.2 F. sp. [AF-6]
100
79/52
Miami-FL
E. interjectus 100
Ambrosia 100
100 Gainesville-FL
Fusarium
100 F. sp. [AF-7]
Clade
9.0 Mya Australia
3.4 Mya
1.4−18.4 0.14−7.1 12.4 Mya
5.6−19.6
F. sp. [AF-8]
100
E. validus
Miami-FL 100
97
Clade A PA-OH-VA-MD
19.3 Mya
24.0 Mya F.sp. [AF-9] 10.7−28.2
100
13.9−34.5 5.1 Mya
Miami-FL / Costa Rica
0.2−10.9
Euwallacea destruens
22468 F. neocosmosporiellum
Papua New Guinea
outgroup outgroup
43467 F. neocosmosporiellum Wallacellus similis
Papua New Guinea
Fig. 1. (Left) One of nine most-parsimonious trees (MPTs) inferred from the combined four-gene data set (EF-1a, ITS + 28S rDNA, RPB1 and RPB2) for 12 Ambrosia Fusarium
Clade (identified by red internode) species [AF-1 through AF-12]. Thickened black internodes identify Clades A and B within the Ambrosia Fusarium Clade. Color-coding of
auxiliary lines are used to link nine fusaria with the Euwallacea ambrosia beetles that cultivate them. The putative fungus-farming xyleborine associated with AF-5 from
Malaysia, AF-9 from Miami-Dade County, Florida and AF-10 from Singapore is unknown. (Right) Maximum parsimony analysis of a six-gene data set (ArgK, CAD, CO1, EF-1a,
16S rDNA and 28S rDNA) resolved seven fungus-farming Euwallacea spp. Two Euwallacea spp., Euwallacea sp. #2 infesting avocado in Miami-Dade County, Florida and
Euwallacea sp. #4 from Chinese tea in Sri Lanka, cultivate two closely related fusaria. The Euwallacea fornicatus Clade, which comprises six species lineages (Euwallacea sp. #1–
6), is identified by a thickened black internode. Maximum likelihood (ML) and MP bootstrap (BS) values P 70%, based on 1000 pseudoreplicates of the data, are indicated
above nodes (ML-BS/MP-BS). Only the ML-BS value is shown if it differed by 65% of the MP-BS value. The Fusarium and Euwallacea phylograms were rooted on sequences of F.
neocosmosporiellum (formerly known as Neocosmospora vasinfecta) and E. destruens + Wallacellus similis, respectively, based on more inclusive analyses (Cognato et al., 2011;
O’Donnell et al., 2013). Divergence times were estimated with BEAST 2.1.3 (Bouckaert et al., 2014) using a birth–death model as the tree prior and an uncorrelated lognormal
relaxed molecular clock. Each divergence time, including the median age (in bold) in millions of years before present (Mya) and the Bayesian 95% highest probability density
interval, was taken from the chronograms presented in Supplementary Figs. S1 and S2. Photos at the top of the tanglegram illustrate club-shaped conidia of Fusarium sp. AF-7
NRRL 62610 from Australia and Euwallacea sp. #1 from Israel. CI, consistency index; MPTs, most-parsimonious trees; PIC, parsimony informative characters; RI, retention
index. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)
15 November 2014
Table 1
Histories of xyleborines included in the phylogenetic analyses (Figs. 1 and 2).
a b
Sequence ID Collection # Host/Substrate Geographic origin
Euwallacea ‘fornicatus’ Zool. Scripta 40:174–186 (2010) Artocarpus communis (breadfruit) Papua New Guinea
Euwallacea destruens Zool. Scripta 40:174–186 (2010) Pterocymbium beccarii (amberoi) Papua New Guinea
Euwallacea interjectus CB-8965 Acer negundo (box elder) Gainesville-FL
Euwallacea validus MK-MD1-1 Ailanthus altissima (tree-of-heaven) MD-USA
Euwallacea validus MK-P1 Ailanthus altissima (tree-of-heaven) OH-USA
Euwallacea validus MK-C1 Ailanthus altissima (tree-of-heaven) PA-USA
Euwallacea sp. #1 SF-1 Persea americana (avocado) Israel
Euwallacea sp. #1 SF-M1 Persea americana (avocado) Israel
Euwallacea sp. #1 AE-55 Quercus robur (English oak) Los Angeles-CA
Euwallacea sp. #1 AC-1 Acer negundo (box elder) Los Angeles-CA
Euwallacea sp. #2 JK-35 Persea americana (avocado) Miami-Dade Co.-FL
Euwallacea sp. #2 JK-37 Persea americana (avocado) Miami-Dade Co.-FL
Euwallacea sp. #2 DC-2 Persea americana (avocado) Miami-Dade Co.-FL
Euwallacea sp. #2 RP-29 Persea americana (avocado) Miami-Dade Co.-FL
E. validus – Euwallacea sp. #2 hybrids RP-30 Persea americana (avocado) Miami-Dade Co.-FL
E. validus – Euwallacea sp. #2 hybrids RP-31 Persea americana (avocado) Miami-Dade Co.-FL
Euwallacea sp. #3 AG-1 Persea americana (avocado) Queensland, Australia
Euwallacea sp. #3 AG-2 Persea americana (avocado) Queensland, Australia
Euwallacea sp. #4 PL-KT2 Camellia sinensis (tea) Kandy, Sri lanka
Euwallacea sp. #4 PL-KT-10 Camellia sinensis (tea) Kandy, Sri lanka
Euwallacea sp. #4 PL-TT1 Camellia sinensis (tea) Talawakelle, Sri Lanka
Euwallacea sp. #5 PN-RW1 Salix laevigata (red willow) San Diego Co.-CA
Euwallacea sp. #5 PN-SY1 Platanus racemosa (California sycamore) San Diego Co.-CA
Wallacellus similis Zool. Scripta 40:174–186 (2010) Pterocymbium beccarii (amberoi) Papua New Guinea
a
The dataset consisted of portions of the following six genes: argK, EF-1a, CAD, cox1, 16S rDNA and 28S rDNA. Euwallacea ‘fornicatus’ was put in single quotes to indicate
that it’s taxonomic identity remains to be determined. Six species level lineages within the morphospecies Euwallacea fornicatus are identified by #1–6. Euwallacea ‘fornicatus’
was designated Euwallacea sp. #6 herein.
b
Sequences of three xyleborines from Papua New Guinea reported on in Zool. Scripta 40:174–186(2010) were downloaded from GenBank. The following two letter code
was used to identify the collectors: AC, Allard Cossé, ARS-USDA, Peoria, IL; AE, Akif Eskalen, U. California–Riverside; AG, Andrew Geering, U. of Queensland, Australia; CB, Craig
Bateman, U. of Florida-Gainesville; DC, Daniel Carrillo, U. Florida-Homestead; MK, Matthew Kasson, U. West Virginia, Morgantown; PL, Pradeepa Liyanage, U. Colombo, Sri
Lanka; PN, Pat Nolan, County of San Diego, Plant Pathology Lab, San Diego; RP, Randy Ploetz, U. Florida-Homestead; SF, Stanley Freeman, fInstitute of Plant Protection, Bet
Dagan, Israel.
323 with PAUP⁄ 4.0b10, employing 1000 pseudoreplicates of the data, cost values: cospeciation = 0, duplication = 1, host switch = 1, line- 353
324 10 random taxon-addition sequences per replicate, TBR branch age sorting = 1, and failure to diverge = 1. The Euwallacea and 354
325 swapping, with MAXTREES set to automatically increase by 100. Fusarium phylogenies were inferred, respectively, from datasets 355
326 Maximum likelihood (ML) analyses were conducted with GARLI that consisted of sequences of the nine Euwallacea and fusaria that 356
327 ver. 1.0 (ML, Zwickl, 2006) on the CIPRES Science Gateway site were linked by the auxiliary lines in Fig. 1. Because predefined 357
328 (http://www.phylo.org) using the default GTR + C + I model of event costs are difficult to predict a priori, the costscape tool in 358
329 molecular evolution. ML bootstrap support was assessed by 1000 the Xscape cophylogeny software tool set (Libeskind-Hadas et al., 359
330 pseudoreplicates of the data. 2014; http://www.cs.hmc.edu/~hadas/xscape/) was used to com- 360
331 BEAST ver. 2.1.3 (Bouckaert et al., 2014) was used to generate pute the set of Pareto-optimal event count vectors and visualize 361
332 divergence-time estimates for Euwallacea and Fusarium as the costs of host switch or loss over the range of 0.2–5 in the 362
333 described in O’Donnell et al. (2013). Published stem and crown Euwallacea–Fusarium dataset. The algorithmic approach used in 363
334 ages of the Xyleborini (Jordal and Cognato, 2012) and dates for five the costscape tool in Xscape uses a reconciliation algorithm that 364
335 ascomycetous fungi (Kasson et al., 2013) were used to obtain a pri- can in theory result in temporal inconsistencies. By contrast, the 365
336 ori divergence dates for the BEAST analyses. Separate uncorrelated one used in Jane 4 does not permit such inconsistencies, but it also 366
337 lognormal relaxed molecular clock analyses were conducted using does not guarantee that all optimal solutions will be found. Empir- 367
338 birth–death (Gernhard, 2008) and calibrated Yule models as the ical p-values were computed using the sigscape tool in Xscape for 368
339 tree priors. An uncorrelated lognormal relaxed molecular clock host switch and loss costs ranging from 1 to 5 based on 1000 369
340 (Drummond and Rambaut, 2007) was employed to compensate permutations. 370
341 for potential differences in clade-specific rates. TRACER ver. 1.6
342 (Rambaut and Drummond, 2007) was used to determine that the 3. Results 371
343 Markov Chain Monte Carlo (MCMC) had reached convergence
344 and Tree Annotator ver. 1.8.0 was used to summarize the results 3.1. Phylogenetic diversity of the Fusarium–Euwallacea mutualism 372
345 as maximum clade credibility (MCC) trees (Drummond and
346 Rambaut, 2007; Supplementary Figs. S1 and S2). Statistical uncer- By DNA typing Fusarium, and the Euwallacea from which they 373
347 tainty in the divergence-time estimates was assessed by Bayesian were isolated, we were able to obtain a preliminary estimate of 374
348 95% highest probability density (HPD) intervals. symbiont fidelity, as reflected by the color-coded auxiliary lines 375
used to link nine fusaria with the seven Euwallacea spp. that culti- 376
349 2.4. Cophylogenetic analysis of the Euwallacea–Fusarium mutualism vate them (Fig. 1). For the AFC fusaria, we continue to follow the 377
previously developed notation for species lineages (Kasson et al., 378
350 Jane 4 (Conow et al., 2010) and CoRe-PA (Merkle et al., 2010), 2013), which now number twelve (AF 1–12), with AF-1 and AF-2 379
351 two parsimony-based software tools, were used to reconstruct representing F. ambrosium and F. euwallaceae, respectively (Fig. 1). 380
352 cophylogenetic scenarios, using the following predefined event MP phylogenetic analysis of a 25-taxon, 4-gene dataset (EF-1a, 381
15 November 2014
contributed similar phylogenetic signal, and together 80% of the 390

PIC in the combined dataset, RPB2 outperformed RPB1 and the other 391
genes in providing P70% parsimony bootstrap support for nine of 392
the 11 AFC species with more than one isolate. By contrast, phylo- 393
genetic analysis of the ITS + 28S rDNA revealed that it contributed 394
the least phylogenetic signal with only 24 PIC; bootstrap analyses 395
of this region only supported the monophyly of three AFC species 396
(Table 3; Supplementary Table S6). Moreover, bootstrapping the 397
ITS rDNA only resolved three of the 12 AFC species (AF-6, AF-8 398
and AF-9). Although the monophyly of Fusarium sp. AF-10 from Sin- 399
gapore could not be tested because it was represented by a single, 400
genetically divergent isolate, GCPSR-based analyses of the individ- 401
ual and combined dataset supported the genealogical exclusivity of 402
the other 11 AFC lineages (Fig. 1). 403
The combined analysis is presented as the strongest hypothesis 404
of evolutionary relationships within the AFC in that 16 nodes and 405
the 11 species represented by more than one isolate were strongly 406
supported by ML and/or MP bootstrapping (Fig. 1, left; Table 3; 407
Supplementary Table S6). Three putative novel AFC species were 408
discovered in the present study: AF-10 (NRRL 62941 = IMI 409
351954) from an unknown host tree in Singapore, AF-11 (NRRL 410
62943 and 62944) from Chinese tea (Camellia sinensis) in Sri Lanka, 411
and AF-12 (NRRL 62945 and 62946) from California sycamore 412
(Platanus racemosa) and red willow (Salix laevigata Bebb) in San 413
Diego County, California (Table 2). The latter species was resolved 414
as a sister to Fusarium spp. AF-2, AF-3 and AF-4 with modest sup- 415
port, but no clear relationship between AF-10 and AF-11 and the 416
other AFC lineages was inferred. With the addition of these taxa, 417
bootstrap support for the AFC and the two clades that comprise 418
it (Clades A and B; Kasson et al., 2013; Fig. 1) remained very strong. 419
Support for the backbone of AFC Clade B, which includes the TSHB 420
and PSHB associates, F. ambrosium and F. euwallaceae, was weak as 421
evidenced by four nodes that were not supported by ML and MP 422
bootstrapping (Fig. 1). 423
Maximum likelihood and maximum parsimony analyses of a 424
23-taxon, 6-gene Euwallacea dataset (EF-1a, CAD, ArgK, 28S rDNA 425
D2/D3, CO1 and 16S rDNA) that totaled 3.6 kb and contained 499 426
PIC found 180 MPTs 951 steps in length (Fig. 1, right; Table 4). 427
The Euwallacea phylogeny used sequences of E. destruens Blandford 428
and Wallacellus similis (Ferrari) from Papua New Guinea as out- 429
groups, based on more inclusive analyses of the Xyleborini 430
(Cognato et al., 2011). The three morphologically defined Euwalla- 431
Fig. 2. Opposed maximum parsimony (MP) phylograms inferred from the nuclear cea species analyzed resolved as strongly supported monophyletic 432
large subunit 28S rDNA D2/D3 regions (left) and mitochondrial cytochrome oxidase groups (Fig. 1). The clade corresponding to the E. fornicatus mor- 433
subunit 1 (CO1) gene (right) from five Euwallacea spp., three of which are listed as phology, however, was divided into five strongly supported species 434
Euwallacea sp. #1–3. Species are distinguished using the same color codes as in
Fig. 1. Euwallacea sp. #2 is resolved as a genealogically exclusive lineage in the 28S
lineages and one distinct singleton from Papua New Guinea. The 435
rDNA phylogeny, but it is paraphyletic in the CO1 gene tree. In addition to six lineages comprising the E. fornicatus Clade (EFC) are heretofore 436
possessing an orthologous CO1 allele, two Euwallacea sp. #2 collections from referred to as Euwallacea sp. #1–6. The individual genes varied 437
avocado in Miami-Dade County, Florida (RP-30 and RP-31 in bold black font) share considerably in their contribution of phylogenetic signal, with 438
an identical putatively xenologous allele with E. validus. The most plausible
CO1 (187 PIC) and the 28S rDNA D2/D3 regions (154 PIC) together 439
explanation is hybrid introgression between E. validus and Euwallacea sp. #2.
Because these exotic xyleborines are not known to be sympatric in North America, contributing approximately two-thirds of the PIC, whereas EF-1a 440
we speculate that the interspecific hybridization event took place in Asia where with 17 PIC contributed the least. Although the monophyly of a 441
they are endemic. The phylograms were rooted on sequences of Wallacellus similis single collection from Papua New Guinea designated as Euwallacea 442
from Papua New Guinea (Cognato et al., 2011). Numbers above internodes sp. #6 E. ‘fornicatus’ (Cognato et al., 2011) could not be tested, the 443
represent MP bootstrap support based on 1000 pseudoreplicates of the data. CI,
consistency index; MPTs, most-parsimonious trees; PIC, parsimony informative
other seven Euwallacea lineages were supported as genealogically 444
characters; RI, retention index. exclusive by GCPSR-based analyses of the individual and combined 445
dataset (Fig. 1, right). With the exception of the 373 bp EF-1a gene 446
partition that possessed too little phylogenetic signal to support 447
382 ITS + 28S rDNA, RPB1 and RPB2) for the AFC that totaled 4.9 kb and the monophyly of any of the Euwallacea spp., bootstrap analyses 448
383 comprised 388 parsimony informative characters (PIC) recovered of the five other individual partitions supported the monophyly 449
384 nine equally most-parsimonious trees 566 steps in length. MP of one to four of the seven Euwallacea lineages at P70% (Table 4, 450
385 and ML methods resolved the same topology. The AFC ingroup Supplementary Table S7). As expected, bootstrapping the com- 451
386 was rooted using Fusarium neocosmosporiellum O’Donnell and Geis- bined dataset provided the strongest phylogenetic hypothesis, 452
387 er (formerly Neocosmospora vasinfecta E.F. Sm.) as the outgroup given that 11 nodes and the seven Euwallacea lineages represented 453
388 based on more inclusive analyses of the F. solani species complex by more than one collection received strong (97–100%) bootstrap 454
389 (O’Donnell et al., 2008). While RPB1 (159 PIC) and RPB2 (155 PIC) support (Fig. 1, Table 4). 455
15 November 2014
Table 2
Histories of Ambrosia Fusarium Clade isolates included in the phylogenetic analysis (Fig. 1, left).
a b c
Sequence ID NRRL and equivalent # Xyleborine associate Host/Substrate Geographic origin
AF-1 Fusarium ambrosium NRRL 20438 = IMI 296597 Euwallacea ‘fornicatus’ Camellia sinensis (tea) India
AF-1 Fusarium ambrosium NRRL 62942 = KOD 799 = PL-KH10 Euwallacea sp. #4 Camellia sinensis (tea) Sri Lanka
AF-2 Fusarium euwallaceae NRRL 54727 = CBS 135859 Euwallacea sp. #1 Persea americana (avocado) Israel
AF-2 Fusarium euwallaceae NRRL 62626 = AE-1854 Euwallacea sp. #1 Acer negundo (box elder) Los Angeles, CA
AF-3 Fusarium sp. NRRL 62606 = CB-1499 Euwallacea interjectus Acer negundo (box elder) Gainesville-FL
AF-3 Fusarium sp. NRRL 62629 = CB-1191 Euwallacea interjectus Acer negundo (box elder) Gainesville-FL
AF-4 Fusarium sp. NRRL 62578 = FRC S-2576 Euwallacea validus Ailanthus altissima (tree-of-heaven) PA-USA
AF-4 Fusarium sp. NRRL 62579 = FRC S-2581 Euwallacea validus Ailanthus altissima (tree-of-heaven) PA-USA
AF-5 Fusarium sp. NRRL 22231 = IMI 110107 unknown Hevea brasiliensis (rubber tree) Malaysia
AF-5 Fusarium sp. NRRL 46518 = FRC S-2075 unknown Hevea brasiliensis (rubber tree) Malaysia
AF-6 Fusarium sp. NRRL 62590 = RP-AF9 Euwallacea sp. #2 Persea americana (avocado) Miami, FL
AF-6 Fusarium sp. NRRL 62591 = RP-16–1B Euwallacea sp. #2 Persea americana (avocado) Miami, FL
AF-7 Fusarium sp. NRRL 62610 = AG-1 Euwallacea sp. #3 Persea americana (avocado) Queensland, Australia
AF-7 Fusarium sp. NRRL 62611 = AG-2 Euwallacea sp. #3 Persea americana (avocado) Queensland, Australia
AF-8 Fusarium sp. NRRL 62584 = RP-Amb2 Euwallacea sp. #2 Persea americana (avocado) Miami, FL
AF-8 Fusarium sp. NRRL 62585 = RP-AF4 Euwallacea sp. #2 Persea americana (avocado) Miami, FL
AF-9 Fusarium sp. NRRL 22643 = ATCC 44215 Xyleborus ferrugineus na Costa Rica
AF-9 Fusarium sp. NRRL 66088 = KOD 838 unknown Delonix regia (royal poinciana) Miami, FL
AF-10 Fusarium sp. NRRL 62941 = KOD 147 = IMI 351954 unknown unknown Singapore, Malaysia
AF-11 Fusarium sp. NRRL 62943 = KOD 796 = PL-KH1 Euwallacea sp. #4 Camellia sinensis (tea) Sri Lanka
AF-11 Fusarium sp. NRRL 62944 = KOD 797 = PL-KH2 Euwallacea sp. #4 Camellia sinensis (tea) Sri Lanka
AF-12 Fusarium sp. NRRL 62945 = KOD 793 = AE-FD420-G47 Euwallacea sp. #5 Platanus racemosa (California sycamore) San Diego, CA
AF-12 Fusarium sp. NRRL 62946 = KOD 795 = AE-FD422-G69 Euwallacea sp. #5 Platanus racemosa (California sycamore) San Diego, CA
Fusarium neocosmosporiellum NRRL 22468 = CBS 562.70 na Stored peanuts Guinea
Fusarium neocosmosporiellum NRRL 43467 = CBS 139182 na Human eye LA-USA
a
Genealogical concordance phylogenetic species recognition (Taylor et al., 2000) revealed that partial EF-1a can resolve all of the fusaria included in the present study
except for AF-4 and AF-12.
b
AE, Akif Eskalen, U. California–Riverside; AG, Andrew Geering, U. of Queensland, Australia; ATCC, American Type Culture Collection, Manassas, VA; CB, Craig Bateman, U.
of Florida-Gainesville; CBS, Centraalbureau voor Schimmelcultures—Fungal Biodiversity Center, Utrecht, The Netherlands; FRC, Fusarium Research Center, The Pennsylvania
State University, State College, PA; IMI, CABI Biosciences, Egham, Surrey, England; KOD, maintained in laboratory of Kerry O’Donnell, NCAUR, Peoria, IL; NRRL, ARS Culture
Collection, Peoria, IL; PL, Pradeepa Liyanage, U. Colombo, Sri Lanka; RP, Randy Ploetz, U. Florida-Homestead.
c
Several of the Euwallacea spp. Sampled fit the morphological description of E. fornicatus, but the identity of the latter remains to be determined, given that type material
does not exist. In addition, NRRL 22643 = ATCC 44215 was reported to have been isolated from Xyleborus ferrugineus, but the identification of the beetle is questionable.
Table 3
Statistics for the individual and combined Fusarium partitions.
Locus # bp # MPTsa MPT length CIb RIc UICd PICe # Nodes/# species receiving P70% MP/ML-BSf
EF-1a 678 2 73 0.82 0.91 7 50 8/6
ITS + LSU rDNA 1004 36 38 0.89 0.93 8 24 4/3
RPB1 1588 4 200 0.90 0.94 10 159 11/7
RPB2 1635 366 227 0.74 0.85 8 155 10/9
Combined 4914 9 566 0.78 0.87 33 388 16/11
a
MPTs, most-parsimonious trees.
b
CI, consistency index.
c
RI, retention index.
d
UIC, parsimony uninformative character.
e
PIC, parsimony informative character.
f
Ones that received P70% maximum parsimony (MP) and/or maximum likelihood (ML) bootstrap support.
Table 4
Statistics for individual and combined Euwallacea partitions.
Locus # bp # MPTsa MPT length CIb RIc UICd PICe # Nodes/# species receiving P70% MP/ML-BSf
ArgK 669 >50,000 82 0.89 0.94 23 46 5/2
CAD 469 >50,000 64 0.89 0.93 20 35 3/2
EF-1a 373 >50,000 37 0.89 0.92 16 17 1/0
LSU rDNA 1004 234 282 0.93 0.97 50 154 8/2
CO1 689 15,702 382 0.65 0.85 13 187 7/7
16S rDNA 420 >50,000 93 0.85 0.93 2 60 4/4
Combined 3624 180 951 0.79 0.90 124 499 11/7
a
MPTs, most-parsimonious trees.
b
CI, consistency index.
c
RI, retention index.
d
UIC, parsimony uninformative character.
e
PIC, parsimony informative character.
f
Ones that received P70% maximum parsimony (MP) and/or maximum likelihood (ML) bootstrap support.
15 November 2014
456 Except for the branching order of Euwallacea spp. #2, #3 and #4, lineages were dated to the mid-Miocene and radiation of the most 520
457 which was unresolved, the other nodes in the Euwallacea phylog- recent common ancestor of the remaining ingroup to the late Mio- 521
458 eny received strong ML and MP bootstrap support (Fig. 1). cene 6.5 Mya [95% HPD interval: 2.8–11.2 Mya]. All subsequent 522
459 Although Euwallacea spp. #1–5 and a specimen from Papua New cladogenetic events were dated to the Pliocene 2.3–3.2 Mya and 523
460 Guinea designated Euwallacea sp. #6 matched E. fornicatus pheno- Pleistocene 1.5 Mya [95% HPD interval: 0.5–3.1 Mya]. 524
461 typically (Cognato et al., 2011; NPAG, 2013), the present results
462 suggest that they represent six phylogenetically distinct species.
3.3. Cophylogenetic analyses of the Euwallacea–Fusarium mutualism 525
463 With the exception of Euwallacea sp. #6 that was missing CO1,
464 the other E. fornicatus-like taxa were separated from one another
To explore the relative roles of cospeciation, host-switching, 526
465 by deep divergences (6.2–14.6%) in the CO1 gene tree (data not
duplication, and lineage sorting on a macroevolutionary scale, 527
466 shown). While four of the Euwallacea species were associated with
cophylogenetic analyses of multilocus DNA sequence data from 528
467 a single AFC species, AF-6 and AF-8 were recovered from Euwalla-
nine Euwallacea representing seven species and the nine AFC they 529
468 cea sp. #2 on avocado in Miami-Dade County, Florida, and AF-1 and
cultivate were conducted with two maximum parsimony tree rec- 530
469 AF-11 were associated with Euwallacea sp. #4 on tea in Sri Lanka
onciliation tools, Jane 4 (Conow et al., 2010) and CoRe-PA (Merkle 531
470 (Fig. 1).
et al., 2010). Jane 4 found one solution that had a maximum parsi- 532
471 Interestingly, multilocus sequencing of two Euwallacea sp. #2
mony (i.e., minimum cost) solution of cost 5, with three cospecia- 533
472 individuals (i.e., RP-30 and RP-31) from avocado in Miami-Dade
tions and five host switches (Fig. 3). CoRe-PA analysis resulted in 534
473 County revealed genealogical discordance between trees inferred
the same reconciliation as well as another that differed only in that 535
474 from CO1 and the other loci. PCR assays using conserved and
one of the host shifts was initially to F. ambrosium AF-1 followed by 536
475 Euwallacea sp. #2-specific CO1 primers developed in the present
one to AF-11. CoRe-PA also found a third reconstruction of cost 5, 537
476 study revealed that these two individuals possessed two divergent
but it was determined to be temporally invalid. 538
477 CO1 alleles. Although the partial nuclear 28S rDNA and mitochon-
Since estimating appropriate event costs for maximum parsi- 539
478 drial 16S rDNA placed these collections in Euwallacea sp. #2, as did
mony reconciliation is notoriously difficult, we also used the cost- 540
479 the partial ArgK, CAD and EF-1a sequences, the allele that was pref-
scape tool in the Xscape software toolkit (Libeskind-Hadas et al., 541
480 erentially amplified in these two beetles with the conserved CO1
2014) to explore the effects of a wide range of event costs (cospe- 542
481 PCR primers (Dole et al., 2010; Supplementary Table S3) appeared
ciation cost fixed to 0, duplication normalized to 1, and host switch 543
482 to have been acquired from E. validus or an E. validus-like parent
and loss ranging from 0.2 to 5). The costscape tool reported the 544
483 (Fig. 2). Alignment of the sequences from E. validus and the puta-
number of evolutionary events in all maximum parsimony solu- 545
484 tive E. fornicatus–E. validus hybrids showed they shared an identi-
tions over this cost range as well as the number of distinct optimal 546
485 cal CO1 allele. In silico translations of the putative xenologous CO1
reconciliations. Costscape represents a maximum parsimony solu- 547
486 allele in these two collections suggested that it might be functional
tion with an event count vector of the form <c, d, s, l> indicating 548
487 because the coding region was not interrupted by stop codons or
the number of cospeciations, duplications, host switches, and 549
488 indels. Results of PCR assays specific to the nuclear 28S rDNA,
losses, respectively. One group of solutions had event count vector 550
489 mitochondrial 16S rDNA and CO1 of Euwallacea sp. #2 and E. vali-
<3, 0, 5, 0> and there were two distinct reconstructions in this 551
490 dus indicated that the two putative E. validus – Euwallacea sp. #2
group – both found by CoRePA and one found by Jane (Fig. 3). 552
491 hybrids only possessed orthologous 28S and 16S rDNA sequences,
Another maximum parsimony solution had event count vector 553
492 supporting the interpretation that the E. validus CO1 allele in these
<4, 0, 4, 1>. Three other groups of solutions had event count vectors 554
493 samples was the result of introgressive hybridization rather than
of <4, 2, 2, 1>, <3, 4, 1, 20>, and <3, 5, 0, 27>. However, the latter three 555
494 cross-contamination with E. validus DNA. Furthermore, PCR assay
solutions are only optimal at extreme values for the costs of host 556
495 with Euwallacea sp. #2 CO1-specific PCR primers revealed that
switch or loss event and thus are unlikely to be of interest 557
496 the two hybrid beetles also possessed an orthologous CO1 allele,
(Fig. 4). To assess whether the Euwallacea–Fusarium trees were 558
497 in addition to the xenolog (Fig. 2, Supplementary Table S3).
similar due to chance, a permutation test employing 1000 trials 559
was conducted using the sigscape tool in Xscape. When host switch 560
498 3.2. Divergence-time estimates of the Euwallacea–Fusarium
and loss costs ranging from 1 to 5 were analyzed, 100% of the event 561
499 mutualism
cost space found by sigscape was not significant at the 0.05 level 562
(data not shown). Thus, the null hypothesis that the three cospeci- 563
500 Divergence times for the AFC were estimated with BEAST ver.
ations could have occurred by chance could not be rejected. 564
501 2.1.3 (Bouckaert et al., 2014) using external calibration points as
502 previously described (Kasson et al. 2013; O’Donnell et al., 2013).
503 The Bayesian analysis suggested the AFC evolved in the late 4. Discussion 565
504 Oligocene 24 Mya [95% HPD interval: 13.9–34.5; Supplementary
505 Fig. S1] (Fig. 1, left), followed by a basal split between Clades A and 4.1. Phylogenetic diversity of the Fusarium–Euwallacea mutualism 566
506 B in the middle Miocene 9 Mya [95% HPD interval: 1.4–18.4 Mya].
507 Of the 10 later diverging lineages for which divergence times were In our previous study (Kasson et al., 2013), we documented the 567
508 obtained, seven were dated to the Pliocene 2.3–5.1 Mya and three phylogenetic diversity of fusaria engaged in the mutualistic associ- 568
509 to the Pleistocene 0.6–1.8 Mya (Fig. 1). The poorly resolved back- ation with Euwallacea. In this work we extended the study and dis- 569
510 bone of the AFC Clade B phylogeny, together with the short inter- covered a similar pattern of apparent cryptic speciation among the 570
511 nodes supporting the 10 fusaria within it, suggest they underwent ambrosia beetles that farm them. In Fusarium, we have now 571
512 a relatively rapid diversification over the past 5 million years. detected 11 species according to the criteria of genealogical con- 572
513 Bayesian-derived divergence time estimates were obtained for cordance and non-discordance (GCPSR, Taylor et al. 2000; 573
514 Euwallacea and related Xyleborini with BEAST ver. 2.1.3, using Dettman et al., 2003). These 11 reciprocally monophyletic species 574
515 fossil calibrations previously reported for the Scolytinae (Jordal lineages were supported by bootstrap analyses of one or more of 575
516 and Cognato, 2012). Minimum (crown) time estimates suggest the single-gene genealogies and the combined dataset. Within 576
517 the AFC-farming Euwallacea diverged in the early Miocene 19.3 the sampled Euwallacea, we detected seven phylogenetically dis- 577
518 Mya [95% HPD interval: 10.7–28.2 Mya; Supplementary Fig. S2] tinct species lineages using the same criteria, five of them within 578
519 (Fig. 1, right). The early diverging E. validus and E. interjectus the morphological concept of E. fornicatus. Although the taxonomic 579
15 November 2014
100 100
E-1
= Fusarium F. euwallaceae [AF-2]
100
= Euwallacea
= host shift E. interjectus
100
AF-3
= co-divergence
= duplication
E. validus
AF-4
E-5
AF-12
50
E-4-2
AF-11
E-4-1
F. ambrosium [AF-1]
50 100
E-2-2
100 AF-6
E-3
AF-7
E-2-1
AF-8
Fig. 3. Cophylogenetic analysis of the Euwallacea–Fusarium mutualism conducted with Jane 4 (Conow et al., 2010) using the following event costs: cospeciation = 0,
duplication = 1, host switch = 1, lineage sorting = 1 and failure to diverge = 1. The best-fit reconciliation of the Euwallacea and Fusarium trees included five host shifts (i.e.,
switches to different fusarial symbionts) and three cospeciations. The Fusarium and Euwallacea phylogenies were inferred from sequences of one representative of the nine
fusaria and Euwallacea linked by the color-coded auxiliary lines in Fig. 1. The number by each co-divergence or duplication event represents the percentage of solutions in
which the Euwallacea tree was mapped to this location on the Fusarium tree. (For interpretation of the references to colour in this figure legend, the reader is referred to the
web version of this article.)
New Guinea (Cognato et al., 2011) could not be evaluated employ- 583
ing GCPSR-based criteria, they very likely represent phylogeneti- 584
cally distinct species based on the level of genetic divergence 585
from their sister clades. Two additional findings support our inter- 586
Transfer cost relative to duplication
pretation of species diversity within the AFC and the Euwallacea 587
spp. that cultivate them (i) unique AFC symbionts are cultivated 588
by each of the seven Euwallacea lineages, suggesting symbiosis- 589
related adaptations towards maintaining vertical transmission, 590
and (ii) the divergence-time estimates indicate the youngest spe- 591
cies pairs we sampled within Fusarium and Euwallacea diverged 592
0.6 and 1.5 Mya, respectively. The three fusaria discovered herein 593
produced distinctive club-shaped macroconidia similar to those 594
produced by six other members of the AFC (see Fig. 3 in Kasson 595
et al., 2013). We posit that these spores represent a symbiosis- 596
related adaptation, analogous to nodules, gongylidia or swollen 597
cells produced by fungi that are cultivated, respectively, by ter- 598
mites (Aanen et al., 2002), leaf cutting ants (Schultz and Brady, 599
2008), and ambrosia beetles associated with Ophiostomatales or 600
Microascales (Massoumi Alamouti et al., 2009). 601
The number of ambrosial Fusarium species was extended from 9 602
Loss cost relative to duplication
to 12 in this study by the discovery of AF-11 – Euwallacea sp. #4 on 603
Fig. 4. Pareto-regions computed by the costscape tool in Xscape (Libeskind-Hadas tea in Sri Lanka, AF-12 – Euwallacea sp. #5 on California sycamore 604
et al., 2014) for the 9-taxon Fusarium and Euwallacea datasets employed in the Jane and red willow in San Diego County, California, and the singleton 605
4 cophylogenetic analysis reported in Fig. 3. The legend identifies four event counts
AF-10 from Singapore. The provenance of the previously identified 606
(i.e., cospeciation = 0, duplication = 1, host-switch = 1 and lineage sorting = 1) for
each of the five color-coded regions. The ‘count’ field for each region indicates the
AF-9 remains uncertain. This taxon (= NRRL 22643 received as 607
number of unique reconciliations found by costscape. The two <3, 0, 5, 0> reconcil- ATCC 44215 F. solani) (Norris, 1980) is a phylogenetically distinct 608
iations were deemed the most-parsimonious solutions by CoRe-PA (Merkle et al., member of the AFC that the ATCC catalog indicates was isolated 609
2010); Jane 4 found one of these (Conow et al., 2010). The two optimal from Xyleborus ferrugineus in Costa Rica. However, because the bee- 610
reconciliations differed only as to whether one of the host shifts was initially to
tle used in the Norris publication was not identified, and Euwalla- 611
AF-11 followed by one to F. ambrosium (AF-1), as illustrated in Fig. 3, or the other
way around. Note that the third and fourth regions in the legend occupy zero (i.e., cea and Xyleborus are not reciprocally monophyletic (Cognato et al., 612
<4, 2, 2, 1>) or nearly zero area (i.e., <3, 4, 1, 20>) and are barely visible in the plot. 2011), the identity of the xyleborine beetles reared in these studies 613
(Norris et al., 1967) remains to be determined. Beetle vouchers 614
were not deposited in the Wisconsin Insect Research Collection 615
580 status of the single isolate of F. sp. AF-5 recovered from a beetle- at the University of Wisconsin–Madison (Daniel Young, pers. 616
581 infested rubber tree (Hevea brasiliensis Müll. Arg.) in Malaysia commun.). An isolate of AF-9 (Table 2) was also recovered in the 617
582 and an ambrosia beetle identified as E. fornicatus from Papua present study from beetle-infested royal poinciana, Delonix regia 618
15 November 2014
619 (Boj. ex Hook.) Raf., in Miami-Dade County, Florida, a known host contrast to the CO1 phylogeny (Fig. 2), the four nuclear genes and 685
620 of an ambrosia beetle reported as E. fornicatus (Thomas, 2012). mitochondrial 16S rDNA support the monophyly of these two spe- 686
621 However, further work is needed to ascertain whether AF-9 is cul- cies. Interspecific introgression as reported here is another poten- 687
622 tivated by the Euwallacea ambrosia beetle reported to infest this tial pitfall of taxon identification using a single gene such as CO1 688
623 host. Future studies are also needed to match AFC species AF-5 (Rubinoff et al., 2006). It is also worth noting that DNA sequence 689
624 from a beetle damaged rubber tree (H. brasiliensis) in Malaysia data from this gene cannot be used for the reliable identification 690
625 and AF-10 received as IMI 351954 Fusarium bugnicourtii Brayford of Fusarium due to the presence multiple paralogs in a number of 691
626 (a later synonym of F. ambrosium; Nirenberg, 1990) from an the species (Gilmore et al., 2009). 692
627 unknown woody host in Singapore. Because paternal inheritance of mitochondria is relatively unu- 693
628 Six of the eight Euwallacea lineages analyzed here fit the mor- sual (Dowling and Secor, 1997), particularly in the case of inbreed- 694
629 phological diagnosis of E. fornicatus, the TSHB from Sri Lanka and ing species, special steps were taken to verify our findings of 695
630 India: (1) Euwallacea sp. #1 from avocado and diverse woody hosts hybrid introgression. First, the thorax + abdomen of the two puta- 696
631 in California and Israel (Eskalen et al., 2012; Mendel et al., 2012), tive E. validus – Euwallacea sp. #2 hybrid beetles (RP-30 and RP-31) 697
632 (2) Euwallacea sp. #2 from avocado in Miami-Dade County, Florida, were received in separate tubes filled with 100% ethanol from R. 698
633 (3) Euwallacea sp. #3 from avocado in Queensland, Australia, (4) Ploetz’s lab where E. validus has never been processed. The xenol- 699
634 Euwallacea sp. #4 from Chinese tea in Sri Lanka, (5) Euwallacea ogous origin of the E. validus CO1 allele in two Euwallacea sp. #2 700
635 sp. #5 from red willow and California sycamore in San Diego individuals from Miami-Dade County, Florida was confirmed by 701
636 County, California, and (6) ‘E. fornicatus’ from Papua New Guinea, novel PCR assays specific to the nuclear 28S rDNA, mitochondrial 702
637 in which the fungal symbiont was not identified (Cognato et al., 16S rDNA and CO1 from the two ambrosia beetle species. These 703
638 2011). The taxonomic identity of the actual E. fornicatus is cur- assays revealed that the two putative hybrids only possessed 704
639 rently unclear, as the type specimen has been lost from the orthologous 28S and 16S rDNA sequences, supporting the interpre- 705
640 museum where it was originally deposited. The morphological tation that the E. validus CO1 allele in these samples was the result 706
641 crypsis observed here for Euwallacea is consistent with that of introgressive hybridization rather than cross-contamination 707
642 reported for numerous other Xyleborini, and may be explained with E. validus DNA. We were also able to establish, via a PCR assay 708
643 by lack of strong selection on secondary sexual characters with Euwallacea sp. #2 CO1-specific PCR primers, that the two 709
644 (Farrell et al., 2001; Jordal et al., 2000) and by the relatively recent hybrid individuals possess an orthologous allele. Even though the 710
645 radiation of the six E. fornicatus-like taxa over the past 6.5 Ma putative xenologous CO1 allele of the hybrids was not interrupted 711
646 (Fig. 1, right). Consistent with their obligatory sib-mating and by stop codons and indels (i.e., cryptic sensu Bertheau et al., 2011), 712
647 haplodiploidy, multilocus phylogenetic analyses revealed little suggesting that it might be functional, additional studies are 713
648 allelic variation within the seven highly inbred Euwallacea spp. needed to determine whether it represents a pseudogene that inte- 714
649 we sampled. grated into the mitochondrial or nuclear genomes of Euwallacea sp. 715
650 Results of the present study highlight the need for more infor- #2, or both. Assuming the E. validus CO1 allele is present within the 716
651 mative loci for species-level studies within Euwallacea and the mitochondria, then the paternally and maternally inherited mito- 717
652 AFC. While EF1-a, CAD and ArgK have phylogenetic utility at higher chondrial genomes may have recombined because the 16S mito- 718
653 taxonomic levels within various Coleoptera (Cognato et al., 2011; chondrial rDNA was inherited from Euwallacea sp. #2. Efforts are 719
654 Wild and Maddison, 2008), our results indicate that these three being made to obtain additional samples of the putative E. validus 720
655 genes have very limited utility for species-level studies of Euwall- – Euwallacea sp. #2 hybrid so that it can be subjected to a whole 721
656 acea that cultivate AFC. Similarly, of the four unlinked genes genome analysis. 722
657 sequenced in the AFC, the fungal barcode locus ITS rDNA and D1/ Paternal inheritance of mitochondrial DNA has been reported in 723
658 D2 region of the 28S rDNA contained the least phylogenetic signal diverse insects (Arunkumar et al., 2006; Meusel and Moritz, 1993; 724
659 in that it only contributed 24 phylogenetically informative charac- Sherengul et al., 2006; Wolff et al., 2013; Zakharov et al., 2009), 725
660 ters. The availability of a whole genome sequence of F. euwallaceae other animals including humans (Bromham et al., 2003; Kvist 726
661 NRRL 62626 (Stajich and O’Donnell, unpubl.), and one of the clo- et al., 2003; Zhao et al., 2004; Zouros et al., 1994) and plants 727
662 sely related pea pathogen, F. solani (Mart.) Sacc. f. sp. pisi W.C. Sny- (Neale et al., 1989). Given the well-documented cases of introgres- 728
663 der and H.N. Hansen (Coleman et al., 2009), should provide a sion of mtDNA (Ballard, 2000; Jordal and Kambestad, 2013; Shaw, 729
664 wealth of phylogenetically informative loci for species-level stud- 2002), and Wolbachia infections in diverse arthropods (Werren 730
665 ies within the AFC. et al., 2008), including ones we detected in E. validus and Euwalla- 731
666 Preliminary evidence was obtained consistent with hybrid intro- cea sp. #2 (O’Donnell, unpubl.) and other xyleborine beetles 732
667 gression and parental leakage of the CO1 gene of E. validus into the (Kawasaki et al., 2010), it is worth determining whether the puta- 733
668 population of Euwallacea sp. #2 in Miami-Dade County, Florida tive hybrid introgression of mitochondrial DNA we discovered was 734
669 based on analyses of the individual nuclear and mitochondrial gene driven by Wolbachia-infected Euwallacea (Hurst and Jiggins, 2004). 735
670 partitions. The xenologous allele was preferentially amplified with Of the 11 independent evolutionary origins of fungus-farming 736
671 conserved CO1 primers (Dole et al., 2010), whereas the orthologous that have been documented in the weevil subfamily Scolytinae, 737
672 copy was only detected using Euwallacea sp. #2-specific CO1 prim- the bark and ambrosia beetles (Jordal and Cognato, 2012), only 738
673 ers developed in the present study (Supplementary Table S3). If one appears to involve the fungiculture of Fusarium – that observed 739
674 these two species did hybridize, it probably took place in Asia in the xyleborine genus Euwallacea. All other known ambrosia bee- 740
675 because they do not appear to be sympatric in North America. tles engage in mutualism with symbionts that belong to other fun- 741
676 Moreover, the hybrid introgression is assumed to have taken place gal genera, mostly from Ophiostomatales and Microascales 742
677 relatively recently given that the Euwallacea sp. #2 CO1 xenolog (Massoumi Alamouti et al., 2009). The available data suggest that 743
678 and E. validus ortholog are identical (Fig. 2). It is worth mentioning this mutualistic symbiosis arose monophyletically in the late Oli- 744
679 that the haploid, blind and flightless males occasionally leave their gocene to early Miocene 19–24 Mya (Kasson et al., 2013; present 745
680 galleries and wander around trees in search of other galleries, thus study). Euwallacea is a derived genus within Xyleborini, and pos- 746
681 providing the opportunity for interspecific matings (Hulcr, sesses the ancestral mandibular mycangium (Cognato et al., 747
682 unpubl.). We discounted incomplete lineage sorting because (i) E. 2011). Therefore its ancestors probably already engaged in the 748
683 validus and Euwallacea sp. #2 are distantly related, having shared ambrosia symbiosis, but with different fungi. The observation of 749
684 a most recent common ancestor 12.4 Mya (Fig. 1), and (ii) in Raffaelea (Ophiostomatales) and Graphium (Microascales) spp. in 750
15 November 2014
751 association with E. validus (Kasson et al., 2013) may reflect a partial of Euwallacea spp. or their galleries suggests that they may be obli- 817
752 retention of this ancestral character. gate symbionts. 818
753 Prior to the outbreak of Fusarium dieback in California (Eskalen Interestingly, once Euwallacea galleries are abandoned they rap- 819
754 et al., 2013) and Israel (Mendel et al., 2012), the best studied idly become overgrown by conidial fungi, including fast growing 820
755 example of Fusarium fungiculture was the tea shot hole borer members of the F. solani species complex, which are otherwise 821
756 (TSHB), E. fornicatus (formerly known as Xyleborus fornicatus), common endophytes and saprophytes on living and dead woody 822
757 which cultivates F. ambrosium in Chinese tea in Sri Lanka and India plant hosts. Therefore, laboratory rearing experiments may help 823
758 (Brayford 1987; Gadd and Loos, 1947). Fusaria were noticeably elucidate how the health of active fungus gardens is maintained. 824
759 absent, however, from two key prior phylogenetic analyses of Research is also needed to ascertain whether Euwallacea spp., as 825
760 ambrosia fungi (Massoumi Alamouti et al., 2009; Farrell et al., reported for X. ferrugineus (Kok et al., 1970), can complete their life 826
761 2001) that included mostly ophiostomatoid and to a lesser extent cycles on a fungus-free medium when fed sterols such as ergos- 827
762 microascalean fungi. Fusaria have been recovered from numerous terol isolated from the Fusarium they cultivate. The successful lab- 828
763 phylogenetically diverse insects (O’Donnell et al., 2012), including oratory rearing of Euwallacea spp. #1 and #2 on artificial media 829
764 wood-boring beetles, but the nature of these associations are (Carrillo and Cossé, unpubl.) should facilitate experiments directed 830
765 largely unknown. The Asian longhorned beetle, Anoplophora at elucidating basic aspects of their fungiculture, behavior and 831
766 glabripennis Motschulsky (Geib et al., 2008) yielded a lignocellulose chemical ecology that are not feasible in their woody hosts 832
767 degrading member of the F. solani species complex (FSSC): (Biedermann et al., 2009). 833
768 phylogenetic species FSSC 6 and the xyleborine ambrosia beetle
769 X. ferrugineus from Costa Rica was reportedly associated with 4.2. Divergence-time estimates of the Euwallacea–Fusarium 834
770 Fusarium sp. AF-9 NRRL 22643 (Norris, 1980). Other fusaria have mutualism 835
771 been reported from various other ambrosia beetles (Kolarik and
772 Hulcr, 2008; Kostovcik et al., 2014), but their phylogenetic identity Our analyses suggest a remarkable temporal correlation 836
773 and ecological function remain unknown. In contrast to the between the origin of the AFC-farming Euwallacea and the diversi- 837
774 Fusarium–Euwallacea mutualism, where the Fusarium symbiont is fication of the AFC clade. Dated scolytine fossils (Jordal and 838
775 transmitted in mycangia and cultivated by females in galleries as Cognato, 2012) were used to date the evolutionary origin of the 839
776 a source of nutrition, A. glabripennis vertically transmits FSSC 6 in AFC-farming Euwallacea to the early Miocene 19.3 Mya [95% HPD 840
777 their gut (Geib et al., 2012), whereas various other fusaria associ- interval: 10.7–28.2], which suggests the clade evolved early within 841
778 ated with other clades of ambrosia beetles are mostly vectored the spectacular radiation of the Xyleborini 21 Mya (Jordal and 842
779 phoretically (Bateman and Hulcr, unpubl.). Cognato, 2012; Jordal et al., 2000). In the absence of a fossil record 843
780 In addition to sib-mating and haplodiploidy, the rapid radiation for Fusarium, five external calibration points were used to date the 844
781 of the Xyleborini has been attributed to their specialization on evolutionary origin of the AFC as previously described (Kasson 845
782 ambrosia fungi that concentrate nutrients from living xylem tissue, et al., 2013; O’Donnell et al., 2013). Based on the coincident diver- 846
783 thus allowing the beetles to greatly increase their range of woody sification-time estimates and their high endemism in southern 847
784 hosts (Jordal et al., 2000). This is dramatically illustrated by the Asia, the available data suggests the Euwallacea–Fusarium mutual- 848
785 broad host range of the F. euwallaceae – Euwallacea sp. #1 mutualism may have first evolved near the Oligocene–Miocene boundary 849
786 ists in southern California (Eskalen et al., 2013). While specificity to 19–24 Mya in tropical or subtropical Asia. The dated phylogeny 850
787 tree hosts is typically low among ambrosia beetles, specificity to indicates that the majority of cladogenetic events within the AFC 851
788 their fungal symbionts may be high. Five different Euwallacea appear to have taken place relatively rapidly over the past 5 Mya, 852
789 spp. analyzed here each cultivates a single AFC species. Only which helps explain the short internodes supporting the terminal 853
790 Euwallacea sp. #2 on avocado from Miami-Dade County, Florida taxa and poor support for the inferred cladogenetic events within 854
791 and Euwallacea sp. #4 from Chinese tea in Sri Lanka cultivate two Clade B. 855
792 different fusaria, but this is still a relatively narrow range of sym- The Euwallacea–Fusarium symbiosis is relatively young com- 856
793 bionts compared to other ambrosia beetles, such as Xyleborus spp., pared to fungiculture in other ambrosia beetle groups that employ 857
794 where a single beetle can carry multiple symbiont genera (Carrillo ophiostomoid and microascalean fungi, some of which date to more 858
795 et al., 2013; Kostovcik et al., 2014). Although preliminary data sug- than 40 Mya (Jordal and Cognato, 2013). However, within 859
796 gest that Euwallacea sp. #1 from Israel and California can complete Xyleborini, Euwallacea and its association with Fusarium appears 860
797 its life cycle when grown on F. euwallaceae (AF-2), but not on F. to have evolved soon after the origin of the whole tribe approxi- 861
798 ambrosium (AF-1) (Freeman et al., 2012), additional experiments mately 20 Mya. It is likely that the xyleborine ancestors of Euwalla- 862
799 are needed to determine whether this feed requirement is obligate. cea that were already farming ophiostomatalean and microascalean 863
800 Preliminary data indicate that the culture medium may require fungi were associated with different fungal clades, and the AFC was 864
801 optimization for each Euwallacea species (Cossé, unpubl.). adopted by Euwallacea de novo. This symbiosis is also younger than 865
802 There are several interesting parallels between the Euwallacea– those of the fungus-farming termites (Nobre et al., 2011) and ant 866
803 Fusarium and other ambrosia beetle – fungus mutualisms, and the agriculture (Mueller et al., 1998; Schultz and Brady, 2008) that 867
804 Macrotermitinae termites – Termitomyces (Aanen et al., 2002) ecto- are estimated to have arisen approximately 31 and 50 Mya, respec- 868
805 symbioses. They all appear to have evolved monophyletically with- tively. Still, the divergence-time estimates should be viewed as rel- 869
806 out any known reversals to a free living state, the fungal symbionts ative chronological events and not as absolute dates, as evidenced 870
807 are transmitted vertically by one parent, host switching is com- by the broad confidence intervals on the dated phylogenies. This 871
808 mon, and during each generation the larvae must acquire a suitable caveat is offered because dated phylogenies are well known to be 872
809 fungal symbiont (Nobre et al., 2011). Symbiont transmission in subject to several sources of error, including rate heterogeneity 873
810 Euwallacea spp. #2 and #4 is mixed, in that fusaria are transmitted among clades and uncertainty in the fossil ages used as calibration 874
811 vertically in mycangia, but horizontal transfers via host-switching points (Berbee and Taylor, 2010; Taylor and Berbee, 2006). 875
812 have occurred (Bright and Bulgheresi, 2010). With the exception of
813 two early diverging taxa within AFC Clade A that are only known 4.3. Cophylogenetic analyses of the Euwallacea–Fusarium mutualism 876
814 from dead wood from Sri Lanka (Kasson et al., 2013), which we
815 speculate might be facultative associates of Euwallacea, the tight In contrast to other methods for comparing phylogenies 877
816 association of later diverging members of the AFC with mycangia (Hughes et al., 2007), we elected to use the maximum parsimony 878
15 November 2014
879 event-based cophylogenetic tools Jane 4 (Conow et al., 2010) and et al., 2013). While our results indicate that some Euwallacea spp. 945
880 CoRe-PA (Merkle et al., 2010) because they can accommodate mul- appear to have switched AFC symbionts in Asia, it remains to be 946
881 tisymbiont hosts and molecular phylogenies that are not fully determined whether these invasive beetles can complete their life 947
882 resolved. These tools use a single set of event costs that are either cycle cultivating newly acquired ambrosia fungi, including other 948
883 specified a priori by the user or automatically inferred based on a members of the AFC. At least two pairs of exotic Euwallacea spp. 949
884 mathematical objective. Since tree reconciliations are known to are already sympatric within the United States: Euwallacea sp. #2 950
885 be sensitive to the choice of event costs, we also used the costscape and E. interjectus are increasingly common throughout Florida (K. 951
886 tool in the Xscape toolkit (http://www.cs.hmc.edu/~hadas/xscape/) Okins, Cooperative Agricultural Pest Survey, Florida Department 952
887 to categorize and count all possible maximum parsimony reconcil- of Agriculture and Consumer Services, pers. comm.) and E. validus 953
888 iations over a broad range of possible event costs (Libeskind-Hadas and E. interjectus have been trapped in northern Georgia and other 954
889 et al., 2014). Of the five groups of potentially optimal solutions that locations throughout the southeastern United States (Atkinson, 955
890 were found by costscape over a 25-fold range of possible event 2014). Now that they are sympatric, the chance that they might 956
891 costs (cospeciation = 0, duplication normalized to 1, switch and cohabit common hosts and exchange AFC symbionts is theoreti- 957
892 loss both varying from 0.2 to 5), one group was found by both Jane cally possible. It is worth noting that the exotic wood-inhabiting 958
893 4 and CoRe-PA with 3 cospeciations, 0 duplications, 5 switches, and wasp Sirex switched Amylostereum mutualists after introduction 959
894 0 losses. There were two nearly identical reconciliations with these into Canada (Wooding et al., 2013). Also, six native or previously 960
895 event counts. Costscape also found a solution with 4 cospeciations, established species of ambrosia beetles began transmitting Raffael- 961
896 0 duplications, 4 switches, and 1 loss that was not found by Jane ea lauricola T.C. Harr., Fraedrich and Aghayeva, the wilt pathogen of 962
897 nor CoRe-PA but this solution is optimal over a wide range of event redbay (Persea borbonia (L.) Spreng.), after it was introduced into 963
898 costs. The remaining reconciliations found by costscape were only the United States from Asia (Carrillo et al., 2014). Similarly, the 964
899 optimal for extreme values of the event costs and were thus unli- seed bug Orsillus maculatus Fieber began vectoring Seiridium cardi- 965
900 kely to be plausible. We used the newly-developed sigscape tool in nale (Wagener) Sutton after this aggressive cypress canker patho- 966
901 Xscape to analyze the statistical significance of the maximum gen was introduced into Europe (Battisti et al., 1999). These and 967
902 parsimony reconciliations over the same broad range of event other examples serve to illustrate that novel pest–pathogen associ- 968
903 costs. This analysis failed to reject the null hypothesis that the ations can pose significant challenges to control and quarantine 969
904 Euwallacea–Fusarium trees might be similar by chance. efforts (Wingfield et al., 2010). Lastly, the host switching that has 970
905 Maximum parsimony-based reconciliation analyses of the helped shaped the Fusarium–Euwallacea mutualism highlights the 971
906 Fusarium–Euwallacea mutualism conducted with Jane 4 and real possibility that more virulent and aggressive combinations 972
907 CoRe-PA suggests the incongruent phylogenies are best explained of these economically destructive mutualists could evolve over 973
908 by multiple host shifts rather than diversifying coevolution (i.e., time. 974
909 host-shift speciation sensu de Vienne et al., 2013). The little evi-
910 dence for cospeciation inferred from the cophylogenetic analysis 5. Uncited references 975
911 is not surprising, given that Euwallacea larvae, and those of other
912 scolytine ambrosia beetles, must acquire fungal symbionts de novo Bensasson et al. (xxxx) and Kok et al. (1979). Q6 976
913 each generation via horizontal transmission, thereby providing the
914 opportunity for host switching (Six, 2012). Indeed, the relationship Acknowledgments 977
915 between many ambrosia beetles and their fungi is much more pro-
916 miscuous than previously thought. Although some beetle groups The authors thank Nathane Orwig for running sequences in the 978
917 have their preferred fungal symbionts (Harrington et al., 2014), NCAUR DNA Core Facility, Thomas White and Ronald Wideman for 979
918 sympatric beetles may also share a common pool of local ambrosia technical assistance in Homestead, CABI Biosciences for providing 980
919 fungi (Carrillo et al., 2013; Kostovcik et al., 2014). Host-shift speci- Fusarium isolates, Paul R. Campbell and Andrew D.W. Geering, The 981
920 ation appears to be the norm not just among beetles, but also in University of Queensland, Australia for collections of Euwallacea 982
921 the broad spectrum of mutualistic associations between other sp. #3, and Pat Nolan, County of San Diego, California for collections 983
922 organisms and fungi (reviewed in de Vienne et al., 2013), including of Euwallacea sp. #5. The mention of firm names or trade products 984
923 the fungus-farming termites (Nobre et al., 2011) and Microbotryum does not imply that they are endorsed or recommended by the US 985
924 anther smuts on caryophyllaceous hosts (Refrégier et al., 2008). The Department of Agriculture over other firms or similar products not 986
925 available data indicate that when host switching did occur in mentioned. The USDA is an equal opportunity provider and 987
926 Euwallacea, Microbotryum and the fungus-growing termites, typi- employer. JH and CB were funded by the USDA Forest Service and Q7 988
927 cally they were to closely related hosts. While the results of these the National Science Foundation, MK was funded by the USDA Forest Q8 989
928 and other studies indicate that long-term cospeciation is rare on a Service’s Forest Health Technology Enterprise Team, and AE and FN 990
929 macroevolutionary scale (de Vienne et al., 2013), the high symbiont were supported by the USDA Forest Service and the California 991
930 fidelity of fungus-growing attine ants in the Cyphomyrmex wheeleri Avocado Commission. 992
931 species group suggests that coevolving lineages can be maintained
932 for several million years (Mehdiabadi et al., 2012).
Appendix A. Supplementary material 993
933 Results of the present study highlight the importance of resolv-
934 ing species limits of invasive pests and pathogens of quarantine
Supplementary data associated with this article can be found, in 994
935 importance. Our GCPSR-based analyses indicate that five different
the online version, at http://dx.doi.org/10.1016/j.fgb.2014.10.014. 995
936 exotic Euwallacea spp. cultivating six different members of the AFC
937 have been introduced into the United States. With the exception of
938 E. validus, first detected on Long Island, New York in 1976 (Rabaglia References 996
939 et al., 2006), the other four Euwallacea spp. appear to have been 997
Aanen, D.K., Eggleton, P., Rouland-Lefévre, C., Guldberg-Frøslev, T., Rosendahl, S.,
940 introduced just within the past one to two decades. The present et al., 2002. The evolution of fungus-growing termites and their mutualistic 998
941 results should help inform quarantine officials and agricultural sci- fungal symbionts. Proc. Natl. Acad. Sci. USA 99, 14887–14892. 999
Arunkumar, K.P., Metta, M., Nagaraju, J., 2006. Molecular phylogeny of silkmoths 1000
942 entists of each species’ genetic diversity, host range and geographic
reveals the origin of domesticated silkmoth, Bombyx mori from Chinese Bombyx 1001
943 distribution so that they can be monitored using molecular mark- mandarina and paternal inheritance of Antheraea proylei mitochondrial DNA. 1002
944 ers such as those developed in this and our previous study (Kasson Mol. Phylogenet. Evol. 40, 419–427. 1003
15 November 2014
1004 Atkinson, T.H., 2014. Bark and Ambrosia Beetles of the US and Canada <http:// Geib, S.M., Scully, E.D., Jimenez-Gasco, M.d.M., Carlson, J.E., Tien, M., 2012. 1090
1005 www.barkbeetles.info/> (accessed 20.08.14). Phylogenetic analysis of Fusarium solani associated with the Asian longhorned 1091
1006 Ballard, J.W.O., 2000. When one is not enough: introgression of mitochondrial DNA beetle, Anoplophora glabripennis. Insects 3, 141–160. 1092
1007 in Drosophila. Mol. Biol. Evol. 17, 1126–1130. Gernhard, T., 2008. The conditioned reconstructed process. J. Theor. Biol. 253, 769– 1093
1008 Battisti, A., Rogues, A., Colombari, F., Frigimelica, G., Guido, A., 1999. Efficient 778. 1094
1009 transmission of an introduced pathogen via an ancient insect-fungus Gilmore, S.R., Gräfenhan, T., Louis-Seize, G., Seifert, K.A., 2009. Multiple copies of 1095
1010 association. Naturwissenschaften 86, 479–483. cytochrome oxidase 1 in species of the fungal genus Fusarium. Mol. Ecol. Res. 9 1096
1011 Beaver, R.A., 1989. Insect–fungus relationships in the bark and ambrosia beetles. In: (Suppl. 1), 90–98. 1097
1012 Wilding, N., Collins, N.M., Hammond, P.M., Webber, J.F. (Eds.), Insect–Fungus Haack, R.A., 2006. Exotic bark- and wood-boring Coleoptera in the United States: 1098
1013 Interactions. Academic Press, London, pp. 121–143. recent establishments and interceptions. Canad. J. For. Res. 36, 269–288. 1099
1014 Bensasson, D., Zhang, D., Hartl, D.L., Hewitt, G.M., Mitochondrial pseudogenes: Harrington, T.C., McNew, D., Mayers, C., Fraedrich, S.W., Reed, S.E., 2014. Ambrosiella 1100
1015 Q9 evolution’s misplaced witnesses. Trends Ecol. Evol. 16, 314–321. roeperi sp. nov. is the mycangial symbiont of the granulate ambrosia beetle, 1101
1016 Berbee, M.L., Taylor, J.W., 2010. Dating the molecular clock in fungi—how close are Xylosandrus crassiusculus. Mycologia 106, 835–845. 1102
1017 we? Fungal Biol. 24, 1–16. Hughes, J., Kennedy, M., Johnson, K.P., Palma, R.L., Page, R.D.M., 2007. Multiple 1103
1018 Bertheau, C., Schuler, H., Krumböck, S., Arthofer, W., Stauffer, C., 2011. Hit or miss in cophylogenetic analyses reveal frequent cospeciation between pelecaniform 1104
1019 phylogeographic analyses: the case of the cryptic NUMTs. Mol. Ecol. Resour. 11, birds and Pectinopygus lice. Syst. Biol. 56, 232–251. 1105
1020 1056–1059. Hulcr, J., Atkinson, T., Cognato, A., Jordal, B., McKenna, D., 2015. Morphology, 1106
1021 Biedermann, P.H.W., Klepzig, K.D., Taborsky, M., 2009. Fungus cultivation by taxonomy and phylogenetics of bark beetles. In: Vega, F., Hofstetter, R. (Eds.), 1107
1022 ambrosia beetles: behavior and laboratory breeding success in three Xyleborine Bark Beetles. Elsevier, Amsterdam. 1108
1023 species. Environ. Entomol. 38, 1096–1105. Hurst, G.D.D., Jiggins, F.M., 2004. Problems with mitochondrial DNA as a marker in 1109
1024 Bouckaert, R., Heled, J., Kühnert, D., Vaughan, T., Wu, C.-H., et al., 2014. BEAST2: a population, phylogeographic and phylogenetic studies: the effects of inherited 1110
1025 software platform for Bayesian evolutionary analysis. PLOS Comput. Biol. 10 symbionts. Proc. Roy. Soc. D 272, 1525–1534. 1111
1026 (e1003537), 2014. http://dx.doi.org/10.1371/journal.pcbi.1003537. Jordal, B.H., Cognato, A.I., 2012. Molecular phylogeny of bark and ambrosia beetles 1112
1027 Brayford, D., 1987. Fusarium bugnicourtii sp. nov., and its relationship to F. tumidum reveals multiple origins of fungus farming during periods of global warming. 1113
1028 and F. tumidum var. coeruleum. Trans. Brit. Mycol. Soc. 89, 347–351. BMC Evol. Biol. 12, 133. 1114
1029 Bright, M., Bulgheresi, S., 2010. A complex journey: transmission of microbial Jordal, B.H., Kambestad, M., 2013. DNA barcoding of bark and ambrosia beetles 1115
1030 symbionts. Nature Rev. 8, 218–230. reveals excessive NUMTs and consistent east–west divergence across Palearctic 1116
1031 Bromham, L., Eyre-Walker, A., Smith, N.H., Maynard Smith, J., 2003. Mitochondrial forests. Mol. Ecol. Res. 14, 7–17. http://dx.doi.org/10.1111/1755-0998.12150. 1117
1032 Steve: paternal inheritance of mitochondria in humans. Trends Ecol. Evol. 18, Jordal, B.H., Normark, B.B., Farrell, B.D., 2000. Evolutionary radiation of an 1118
1033 2–4. inbreeding haplodiploid beetle lineage (Curculionidae, Scolytinae). Biol. J. 1119
1034 Carrillo, D., Duncan, R.E., Peña, J.E., 2012. Ambrosia beetles (Coleoptera: Linnean Soc. 71, 483–499. 1120
1035 Curculionidae: Scolytinae) that breed in avocado wood in Florida. Fla. Kasson, M.T., O’Donnell, K., Rooney, A.P., Sink, S., Ploetz, R.C., et al., 2013. An 1121
1036 Entomol. 95, 573–579. inordinate fondness for Fusarium: Phylogenetic diversity of fusaria cultivated by 1122
1037 Carrillo, D., Duncan, R.E., Ploetz, J.N., Campbell, A.F., Ploetz, R.C., Peña, J.E., 2014. ambrosia beetles in the genus Euwallacea on avocado and other plant hosts. 1123
1038 Lateral transfer of a phytopathogenic symbiont among native and exotic Fungal Gen. Biol. 56, 147–157. 1124
1039 ambrosia beetles. Plant Pathol. 63, 54–62. Kasson, M.T., Short, D.P.G., O’Neal, E.S., Subbarao, K.V., Davis, D.D., 2014. 1125
1040 Cognato, A.I., Hulcr, J., Dole, S.A., Jordal, B.H., 2011. Phylogeny of haplo-diploid, Comparative pathogenicity, biocontrol efficacy, and multilocus sequence 1126
1041 fungus-growing ambrosia beetles (Curculionidae: Scolytinae: Xyleborini) typing of Verticillium nonalfalfae from the invasive Ailanthus altissima and 1127
1042 inferred from molecular and morphological data. Zool. Scripta 40, 174–186. other hosts. Phytopathology 104, 282–292. 1128
1043 Coleman, J.J., Rounsley, S.D., Rodriguez-Carres, M., Kuo, A., Wasmann, C.C., et al., Kawasaki, Y., Ito, M., Miura, K., Kajimura, H., 2010. Superinfection of five Wolbachia 1129
1044 2009. The genome of Nectria haematococca: contribution of supernumerary in the alnus ambrosia beetle, Xylosandrus germanus (Blandford)(Coleoptera: 1130
1045 chromosomes to gene expansion. PLoS Genet. 5, e1000618. Curuculionidae). Bull. Entomol. Res. 100, 231–239. 1131
1046 Conow, C., Fielder, D., Ovadia, Y., Libeskind-Hadas, R., 2010. Jane: a new tool for the Kirkendall, L.R., 1993. Ecology and evolution of biased sex ratios in bark and 1132
1047 cophylogeny reconstruction problem. Algorithms Mol. Biol. 5, 16. http:// ambrosia beetles. In: Wrensch, D.L., Ebbert, M.A. (Eds.), Evolution and Diversity 1133
1048 dx.doi.org/10.1186/1748-7188-5-16. of Sex Ratio: Insects and Mites. Chapman and Hall, New York, pp. 235–345. 1134
1049 Danthanarayana, W., 1968. The distribution and host-range of the shot-hole borer Kirkendall, L.R., Ødegaard, F., 2007. Ongoing invasions of old-growth tropical 1135
1050 (Xyleborus fornicatus Eich.) of tea. Tea Quart. 39, 61–69. forests: establishment of three incestuous beetle species in Central America 1136
1051 de Vienne, D.M., Refrégier, G., López-Villavicencio, M., Tellier, A., Hood, M.E., et al., (Curculionidae, Scolytinae). Zootaxa 1588, 53–62. 1137
1052 2013. Cospeciation vs host-shift speciation: methods for testing, evidence Kok, L.T., Norris, D.M., Chu, H.M., 1979. Sterol metabolism as a basis for a 1138
1053 from natural associations and relation to coevolution. New Phytol. 198, 347– mutualistic symbiosis. Nature 225, 661–662. 1139
1054 385. Kolarik, M., Hulcr, J., 2008. Mycobiota associated with the ambrosia beetle 1140
1055 Dettman, J.R., Jacobson, D.J., Taylor, J.W., 2003. A multilocus genealogical approach Scolytodes unipunctatus (Coleoptera: Curculionidae, Scolytinae). Mycol. Res. 1141
1056 to phylogenetic species recognition in the model eukaryote Neurospora. 113, 44–60. 1142
1057 Evolution 57, 2703–2720. Kostovcik, M., Bateman, C., Stelinski, L.L., Jordal, B.H., Hulcr, J., 2014. The ambrosia 1143
1058 Dole, S.A., Jordal, B.H., Cognato, A.I., 2010. Polyphyly of Xylosandrus Reitter inferred symbiosis is specific in some species and promiscuous in others: evidence from 1144
1059 from nuclear and mitochondrial genes (Coleoptera: Curculionidae: Scolytinae). community pyrosequencing. Int. Soc. Microb. Ecol. J., http://dx.doi.org/10.1038/ 1145
1060 Mol. Phylogenet. Evol. 54, 773–782. ismej.2014.128. 1146
1061 Dowling, T.W., Secor, C.L., 1997. The role of hybridization and introgression in the Kvist, L., Martens, J., Nazarenko, A.A., Orell, M., 2003. Paternal leakage of 1147
1062 diversification of animals. Annu. Rev. Ecol. Syst. 28, 593–619. mitochondrial DNA in the great tit (Porus major). Mol. Biol. Evol. 20, 243– 1148
1063 Drummond, A.J., Rambaut, A., 2007. BEAST: Bayesian evolutionary analysis by 247. 1149
1064 sampling trees. BioMed. Evol. Biol. 7, 214. Libeskind-Hadas, R., Wu, Y.-C., Bansal, M.S., Kellis, M., 2014. Pareto-optimal 1150
1065 Eskalen, A., Stouthamer, R., 2012. Pest Alert: Fusarium dieback on California phylogenetic tree reconciliation. Bioinformatics 30, i86–i95. 1151
1066 avocado trees vectored by tea shot hole borer (Euwallacea fornicatus). University Massoumi Alamouti, S., Tsui, C.K.M., Breuil, C., 2009. Multigene phylogeny of 1152
1067 of California – Riverside <http://ufei.calpoly.edu/forestree/files/collected/ filamentous ambrosia fungi associated with ambrosia and bark beetles. Mycol. 1153
1068 fusarium138734.pdf>. Res. 113, 822–835. 1154
1069 Eskalen, A., Gonzalez, A., Wang, D.H., Twizeyimana, M., Mayorquin, J.S., et al., 2012. Mehdiabadi, N.J., Mueller, U.G., Brady, S.G., Himler, A.G., Schultz, T.R., 2012. 1155
1070 First report of a Fusarium sp. and its vector tea shot hole borer (Euwallacea Symbiont fidelity and the origin of species in fungus-growing ants. Nat. 1156
1071 fornicatus) causing Fusarium dieback on avocado in California. Plant Dis. 96, Commun., http://dx.doi.org/10.1038/ncomms1844. 1157
1072 1070. Mendel, Z., Protasov, A., Sharon, M., Zveibil, A., Ben Yehuda, S., et al., 2012. An Asian 1158
1073 Eskalen, A., Stouthamer, R., Lynch, S.C., Rugman-Jones, P.F., Twizeyimana, M., et al., ambrosia beetle Euwallacea fornicatus and its novel symbiotic fungus Fusarium 1159
1074 2013. Host range of Fusarium dieback and its ambrosia beetle (Coleoptera: sp. pose a serious threat to the Israeli avocado industry. Phytoparasitica 40, 1160
1075 Scolytinae) vector in southern California. Plant Dis. 97, 938–951. 235–238. 1161
1076 Farrell, B.D., Sequeira, A.S., O’Meara, B.C., Normark, B.B., Chung, J.H., et al., 2001. The Merkle, D., Middendorf, M., Wieseke, N., 2010. A parameter-adaptive dynamic 1162
1077 evolution of agriculture in beetles (Curculionidae: Scolytinae and programming approach for inferring cophylogenies. BMC Bioinform. 11 (Suppl. 1163
1078 Platypodinae). Evolution 55, 2011–2027. 1), 560, doi:1186/1471-2105-11-S1-S60. 1164
1079 Freeman, S., Protasov, A., Sharon, M., Mohotti, K., Eliyahu, M., et al., 2012. Obligate Meusel, M.S., Moritz, R.F., 1993. Transfer of paternal mitochondrial DNA during 1165
1080 feed requirement of Fusarium sp. nov., an avocado wilting agent, by the fertilization of honeybee (Apis melifera L.) eggs. Curr. Genet. 24, 539–541. 1166
1081 ambrosia beetle Euwallacea aff. fornicata. Symbiosis 58, 245–251. Mueller, U.G., Rehner, S.A., Schultz, T.R., 1998. The evolution of agriculture in ants. 1167
1082 Freeman, S., Sharon, M., Maymon, M., Mendel, Z., Protasov, A., et al., 2013. Fusarium Science 281, 2034–2038. 1168
1083 euwallaceae sp. nov. – a symbiotic fungus of Euwallacea sp., an invasive Neale, D.B., Marshall, K.A., Sederoff, R.R., 1989. Chloroplast and mitochondrial DNA 1169
1084 ambrosia beetle in Israel and California. Mycologia 105, 1595–1606. are paternally inherited in Sequoia sempervirens D. Don Endl. Proc. Natl. Acad. 1170
1085 Gadd, C.H., Loos, C.A., 1947. The ambrosia fungus of Xyleborus fornicatus Eich. Sci. USA 86, 9347–9349. 1171
1086 Trans. Br. Mycol. Soc. 30, 13–18, plus plate I. Nirenberg, H.I., 1990. Recent advances in Fusarium taxonomy. Stud. Mycol. 32, 91– 1172
1087 Geib, S.M., Filley, T.R., Hatcher, P.G., Hoover, K., Carlson, J.E., et al., 2008. Lignin 101. 1173
1088 degradation in wood-feeding insects. Proc. Natl. Acad. Sci. USA 105, Nobre, T., Koné, N.A., Konaté, S., Linsenmair, K.E., Aanen, D.K., 2011. Dating the 1174
1089 12932–12937. fungus-growing termites’ mutualism shows a mixture between ancient 1175
15 November 2014
1176 codiversification and recent symbiont dispersal across divergent hosts. Mol. Simon, C., Frati, F., Beckenbach, A., Crespi, B., Liu, H., et al., 1994. Evolution, 1219
1177 Ecol. 20, 2619–2627. weighting, and phylogenetic utility of mitochondrial gene sequences and a 1220
1178 Norris, D.M., 1980. Degradation of 14C-labeled lignins and 14C-labeled aromatic compilation of conserved polymerase chain-reaction primers. Ann. Entomol. 1221
1179 acids by Fusarium solani. Appl. Environ. Microbiol. 41, 376–380. Soc. Am. 87, 651–701. 1222
1180 New Pest Advisory Group (NPAG), 2013. Euwallacea fornicatus Eichhoff: Tea shot- Six, D.L., 2012. Ecological and evolutionary determinants of bark beetle-fungus 1223
1181 hole borer <http://www.aphis.usda.gov/plant_health/cphst/npag/downloads/ symbioses. Insects 3, 339–366. 1224
1182 Euwallacea_fornicatusNPAG_LT.pdf>. Swofford, D.L., 2003. PAUP⁄. Phylogenetic analysis using parsimony (⁄and other 1225
1183 O’Donnell, K., Cigelnik, E., Nirenberg, H., 1998. Molecular systematics and methods). Version 4.0b10. Sinauer Associates, Sunderland, Massachusetts. 1226
1184 phylogeography of the Gibberella fujikuroi species complex. Mycologia 90, Taylor, J.W., Berbee, M.L., 2006. Dating divergences in the fungal tree of life: review 1227
1185 465–493. and new analyses. Mycologia 98, 838–849. 1228
1186 O’Donnell, K., Sutton, D.A., Fothergill, A., McCarthy, D., Rinaldi, M.G., et al., 2008. Taylor, J.W., Jacobson, D.J., Kroken, S., Kasuga, T., Geiser, D.M., et al., 2000. 1229
1187 Molecular phylogenetic diversity, multilocus haplotype nomenclature, and Phylogenetic species recognition and species concepts in fungi. Fungal Genet. 1230
1188 in vitro antifungal resistance within the Fusarium solani species complex. J. Clin. Biol. 31, 21–32. 1231
1189 Microbiol. 46, 2477–2490. Thomas, M.C., 2012. Two Asian Ambrosia Beetles Recently Established in Florida. 1232
1190 O’Donnell, K., Humber, R.A., Geiser, D.M., Kang, S., Park, B., et al., 2012. Phylogenetic (Curculionidae: Scolyrinae). Pest Alert. Florida Department of Agriculture and 1233
1191 diversity of insecticolous fusaria inferred from multilocus DNA sequence data Consumer Services, Division of Plant Industry. DACS-P-01647 <http:// 1234
1192 and their molecular identification via FUSARIUM-ID and Fusarium MLST. www.freshfromflorida.com/Divisions-Offices/Plant-Industry/Plant-Industry- 1235
1193 Mycologia 104, 427–445. Publications/Pest-Alerts/Pest-Alert-Two-Asian-Ambrosia-Beetles>. 1236
1194 O’Donnell, K., Rooney, A.P., Proctor, R.H., Brown, D.W., McCormick, S.P., et al., 2013. Werren, J.H., Baldo, L., Clark, M.E., 2008. Wolbachia: master manipulators of 1237
1195 RPB1 and RPB2 phylogeny supports an early Cretaceous origin and a strongly invertebrate biology. Nat. Rev. 6, 741–751. 1238
1196 supported clade comprising all agriculturally and medically important fusaria. Wild, A.L., Maddison, D.R., 2008. Evaluating nuclear protein-coding genes for 1239
1197 Fungal Genet. Biol. 52, 20–31. http://dx.doi.org/10.1016/j.fgb.2012.12.004. phylogenetic utility in beetles. Mol. Phylogenet. Evol. 48, 877–891. 1240
1198 Ploetz, R.C., Hulcr, J., Wingfield, M.J., de Beer, Z.W., 2013. Destructive tree diseases Wingfield, M.J., Slippers, B., Wingfield, B.D., 2010. Novel associations between 1241
1199 associated with ambrosia and bark beetle-associated tree diseases: black swan pathogens, insects and tree species threaten world forests. New Zeal. J. Forest. 1242
1200 events in tree pathology? Plant Dis. 97, 856–872. Sci. 40, S95–S103. 1243
1201 Rabaglia, R.J., Dole, S.A., Cognato, A.I., 2006. Review of American Xyleborina Wolff, J.N., Nafisinia, M., Ballard, J.W.O., 2013. Paternal transmission of 1244
1202 (Coleoptera: Curculionidae: Scolytinae) occurring north of Mexico, with an mitochondrial DNA as an integral part of mitochondrial inheritance in 1245
1203 illustrated key. Ann. Entomol. Soc. Am. 99, 1034–1056. metapopulations of Drosophila simulans. Heredity 110, 57–62. 1246
1204 Refrégier, G., Le Gac, M., Jabbour, F., Widmer, A., Shykoff, J.A., et al., 2008. Wooding, A.L., Wingfield, M.J., Hurley, B.P., Garnas, J.R., de Groot, P., et al., 2013. Lack 1247
1205 Cophylogeny of the anther smut fungi and their caryophyllaceous hosts: of fidelity revealed in an insect-fungal mutualism after invasion. Biol. Lett. 9, 1248
1206 prevalence of host shifts and importance of delimiting parasite species for 20130342. 1249
1207 inferring cospeciation. BMC Evol. Biol. 8, 100. http://dx.doi.org/10.1186/147- Zakharov, E.V., Lobo, N.F., Nowak, C., Hellmann, J.J., 2009. Introgression as a likely 1250
1208 2148-8-100. cause of mtDNA paraphyly in two allopatric skippers (Lepidoptera: 1251
1209 Rubinoff, D., Cameron, S., Will, K., 2006. A genomic perspective on the shortcomings Hesperiidae). Heredity 102, 590–599. 1252
1210 of mitochondrial DNA for ‘‘barcoding’’ identification. J. Hered. 97, 581–594. Zhao, X., Li, N., Guo, W., Hu, X., Liu, Z., Gong, G., Wang, A., Feng, J., Wu, C., 2004. 1253
1211 Schultz, T.R., Brady, S.G., 2008. Major evolutionary transitions in ant agriculture. Further evidence for paternal inheritance of mitochondrial DNA in the sheep 1254
1212 Proc. Natl. Acad. Sci. USA 105, 5435–5440. (Ovis aries). Heredity 93, 399–403. 1255
1213 Shaw, K.L., 2002. Conflict between nuclear and mitochondrial DNA phylogenies Zouros, E., Oberhauser, B.A., Saavedra, C., Freeman, K.R., 1994. An unusual type of 1256
1214 of a recent species radiation: what mtDNA reveals and conceals about modes mitochondrial DNA inheritance in the blue mussel Mytilus. Proc. Natl. Acad. Sci. 1257
1215 of speciation in Hawaiian crickets. Proc. Natl. Acad. Sci. USA 99, 16122– USA 91, 7463–7467. 1258
1216 16127. Zwickl, D.J., 2006. Genetic Algorithm Approaches for the Phylogenetic Analysis of 1259
1217 Sherengul, W., Kondo, R., Matsuura, E.T., 2006. Analysis of paternal transmission of Large Biological Sequence Datasets Under The Maximum Likelihood Criterion 1260
1218 mitochondrial DNA in Drosophila. Genes Genet. Syst. 81, 399–404. [Doctoral Dissertation]. The University of Texas at Austin, p. 115. 1261
1262

O'Donnell Et Al. 2014

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

O'Donnell Et Al. 2014

Uploaded by

Copyright:

Available Formats

YFGBI 2750 No.

of Pages 14, Model 5G

Fungal Genetics and Biology xxx (2014) xxx–xxx

Contents lists available at ScienceDirect

Fungal Genetics and Biology

3 Discordant phylogenies suggest repeated host shifts

2 K. O’Donnell et al. / Fungal Genetics and Biology xxx (2014) xxx–xxx

K. O’Donnell et al. / Fungal Genetics and Biology xxx (2014) xxx–xxx 3

4 K. O’Donnell et al. / Fungal Genetics and Biology xxx (2014) xxx–xxx

4 genes Fusarium Euwallacea 6 genes

steps 1.5 Mya steps

100 F.sp. [AF-5] E. ‘fornicatus’ PNG

3.1 Mya Malaysia Euwallacea

K. O’Donnell et al. / Fungal Genetics and Biology xxx (2014) xxx–xxx 5

6 K. O’Donnell et al. / Fungal Genetics and Biology xxx (2014) xxx–xxx

contributed similar phylogenetic signal, and together 80% of the 390

K. O’Donnell et al. / Fungal Genetics and Biology xxx (2014) xxx–xxx 7

8 K. O’Donnell et al. / Fungal Genetics and Biology xxx (2014) xxx–xxx

K. O’Donnell et al. / Fungal Genetics and Biology xxx (2014) xxx–xxx 9

10 K. O’Donnell et al. / Fungal Genetics and Biology xxx (2014) xxx–xxx

K. O’Donnell et al. / Fungal Genetics and Biology xxx (2014) xxx–xxx 11

12 K. O’Donnell et al. / Fungal Genetics and Biology xxx (2014) xxx–xxx

K. O’Donnell et al. / Fungal Genetics and Biology xxx (2014) xxx–xxx 13

14 K. O’Donnell et al. / Fungal Genetics and Biology xxx (2014) xxx–xxx

You might also like